International Journal of Emerging Technologies in Computational and Applied Sciences by iasir journals

ISSN (ONLINE): 2279-0055 ISSN (PRINT): 2279-0047

Issue 7, Volume 1, 2, 3 & 4 December-2013 to February-2014

International Journal of Emerging Technologies in Computational and Applied Sciences

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

STEM International Scientific Online Media and Publishing House Head Office: 148, Summit Drive, Byron, Georgia-31008, United States. Offices Overseas: India, Australia, Germany, Netherlands, Canada. Website: www.iasir.net, E-mail (s): iasir.journals@iasir.net, iasir.journals@gmail.com, ijetcas@gmail.com

PREFACE We are delighted to welcome you to the seventh issue of the International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS). In recent years, advances in science, technology, engineering, and mathematics have radically expanded the data available to researchers and professionals in a wide variety of domains. This unique combination of theory with data has the potential to have broad impact on educational research and practice. IJETCAS is publishing high-quality, peer-reviewed papers covering topics such as computer science, artificial intelligence, pattern recognition, knowledge engineering, process control theory and applications, distributed systems, computer networks and software engineering, electrical engineering, electric machines modeling and design, control of electric drive systems, non-conventional energy conversion, sensors, electronics, communications, data transmission, energy converters, transducers modeling and design, electro-physics, nanotechnology, and quantum mechanics.

The editorial board of IJETCAS is composed of members of the Teachers & Researchers community who have expertise in a variety of disciplines, including computer science, cognitive science, learning sciences, artificial intelligence, electronics, soft computing, genetic

algorithms,

technology

management,

manufacturing

technology,

electrical

technology, applied mathematics, automatic control , nuclear engineering, computational physics, computational chemistry and other related disciplines of computational and applied sciences. In order to best serve our community, this Journal is available online as well as in hard-copy form. Because of the rapid advances in underlying technologies and the interdisciplinary nature of the field, we believe it is important to provide quality research articles promptly and to the widest possible audience.

We are happy that this Journal has continued to grow and develop. We have made every effort to evaluate and process submissions for reviews, and address queries from authors and the general public promptly. The Journal has strived to reflect the most recent and finest researchers in the field of emerging technologies especially related to computational and applied sciences. This Journal is completely refereed and indexed with major databases like: IndexCopernicus, Computer Science Directory, GetCITED, DOAJ, SSRN, TGDScholar, WorldWideScience, CiteSeerX, CRCnetBASE, Google Scholar, Microsoft Academic Search, INSPEC, ProQuest, ArnetMiner, Base, ChemXSeer, citebase, OpenJ-Gate, eLibrary, SafetyLit, SSRN, VADLO, OpenGrey, EBSCO, ProQuest, UlrichWeb, ISSUU, SPIE Digital Library,

arXiv,

ERIC,

EasyBib,

Infotopia,

WorldCat,

.docstoc

JURN,

Mendeley,

ResearchGate, cogprints, OCLC, iSEEK, Scribd, LOCKSS, CASSI, E-PrintNetwork, intute, and some other databases.

We are grateful to all of the individuals and agencies whose work and support made the Journal's success possible. We want to thank the executive board and core committee members of the IJETCAS for entrusting us with the important job. We are thankful to the members of the IJETCAS editorial board who have contributed energy and time to the Journal with their steadfast support, constructive advice, as well as reviews of submissions. We are deeply indebted to the numerous anonymous reviewers who have contributed expertly evaluations of the submissions to help maintain the quality of the Journal. For this seventh issue, we received 158 research papers and out of which only 82 research papers are published in five volumes as per the reviewersâ&#x20AC;&#x2122; recommendations. We have highest respect to all the authors who have submitted articles to the Journal for their intellectual energy and creativity, and for their dedication to the field of computational and applied sciences.

This issue of the IJETCAS has attracted a large number of authors and researchers across worldwide and would provide an effective platform to all the intellectuals of different streams to put forth their suggestions and ideas which might prove beneficial for the accelerated pace of development of emerging technologies in computational and applied sciences and may open new area for research and development. We hope you will enjoy this seventh issue of the International Journal of Emerging Technologies in Computational and Applied Sciences and are looking forward to hearing your feedback and receiving your contributions.

(Administrative Chief)

(Managing Director)

(Editorial Head)

--------------------------------------------------------------------------------------------------------------------------The International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS), ISSN (Online): 2279-0055, ISSN (Print): 2279-0047 (December-2013 to February-2014, Issue 7, Volume 1, 2, 3 & 4). ---------------------------------------------------------------------------------------------------------------------------

BOARD MEMBERS

        

                 

EDITOR IN CHIEF Prof. (Dr.) Waressara Weerawat, Director of Logistics Innovation Center, Department of Industrial Engineering, Faculty of Engineering, Mahidol University, Thailand. Prof. (Dr.) Yen-Chun Lin, Professor and Chair, Dept. of Computer Science and Information Engineering, Chang Jung Christian University, Kway Jen, Tainan, Taiwan. Divya Sethi, GM Conferencing & VSAT Solutions, Enterprise Services, Bharti Airtel, Gurgaon, India. CHIEF EDITOR (TECHNICAL) Prof. (Dr.) Atul K. Raturi, Head School of Engineering and Physics, Faculty of Science, Technology and Environment, The University of the South Pacific, Laucala campus, Suva, Fiji Islands. Prof. (Dr.) Hadi Suwastio, College of Applied Science, Department of Information Technology, The Sultanate of Oman and Director of IETI-Research Institute-Bandung, Indonesia. Dr. Nitin Jindal, Vice President, Max Coreth, North America Gas & Power Trading, New York, United States. CHIEF EDITOR (GENERAL) Prof. (Dr.) Thanakorn Naenna, Department of Industrial Engineering, Faculty of Engineering, Mahidol University, Thailand. Prof. (Dr.) Jose Francisco Vicent Frances, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Huiyun Liu, Department of Electronic & Electrical Engineering, University College London, Torrington Place, London. ADVISORY BOARD Prof. (Dr.) Kimberly A. Freeman, Professor & Director of Undergraduate Programs, Stetson School of Business and Economics, Mercer University, Macon, Georgia, United States. Prof. (Dr.) Klaus G. Troitzsch, Professor, Institute for IS Research, University of Koblenz-Landau, Germany. Prof. (Dr.) T. Anthony Choi, Professor, Department of Electrical & Computer Engineering, Mercer University, Macon, Georgia, United States. Prof. (Dr.) Fabrizio Gerli, Department of Management, Ca' Foscari University of Venice, Italy. Prof. (Dr.) Jen-Wei Hsieh, Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taiwan. Prof. (Dr.) Jose C. Martinez, Dept. Physical Chemistry, Faculty of Sciences, University of Granada, Spain. Prof. (Dr.) Panayiotis Vafeas, Department of Engineering Sciences, University of Patras, Greece. Prof. (Dr.) Soib Taib, School of Electrical & Electronics Engineering, University Science Malaysia, Malaysia. Prof. (Dr.) Vit Vozenilek, Department of Geoinformatics, Palacky University, Olomouc, Czech Republic. Prof. (Dr.) Sim Kwan Hua, School of Engineering, Computing and Science, Swinburne University of Technology, Sarawak, Malaysia. Prof. (Dr.) Jose Francisco Vicent Frances, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Rafael Ignacio Alvarez Sanchez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Praneel Chand, Ph.D., M.IEEEC/O School of Engineering & Physics Faculty of Science & Technology The University of the South Pacific (USP) Laucala Campus, Private Mail Bag, Suva, Fiji. Prof. (Dr.) Francisco Miguel Martinez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Antonio Zamora Gomez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Leandro Tortosa, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Samir Ananou, Department of Microbiology, Universidad de Granada, Granada, Spain. Dr. Miguel Angel Bautista, Department de Matematica Aplicada y Analisis, Facultad de Matematicas, Universidad de Barcelona, Spain.

           

                  

Prof. (Dr.) Prof. Adam Baharum, School of Mathematical Sciences, University of Universiti Sains, Malaysia, Malaysia. Dr. Cathryn J. Peoples, Faculty of Computing and Engineering, School of Computing and Information Engineering, University of Ulster, Coleraine, Northern Ireland, United Kingdom. Prof. (Dr.) Pavel Lafata, Department of Telecommunication Engineering, Faculty of Electrical Engineering, Czech Technical University in Prague, Prague, 166 27, Czech Republic. Prof. (Dr.) P. Bhanu Prasad, Vision Specialist, Matrix vision GmbH, Germany, Consultant, TIFACCORE for Machine Vision, Advisor, Kelenn Technology, France Advisor, Shubham Automation & Services, Ahmedabad, and Professor of C.S.E, Rajalakshmi Engineering College, India. Prof. (Dr.) Anis Zarrad, Department of Computer Science and Information System, Prince Sultan University, Riyadh, Saudi Arabia. Prof. (Dr.) Mohammed Ali Hussain, Professor, Dept. of Electronics and Computer Engineering, KL University, Green Fields, Vaddeswaram, Andhra Pradesh, India. Dr. Cristiano De Magalhaes Barros, Governo do Estado de Minas Gerais, Brazil. Prof. (Dr.) Md. Rizwan Beg, Professor & Head, Dean, Faculty of Computer Applications, Deptt. of Computer Sc. & Engg. & Information Technology, Integral University Kursi Road, Dasauli, Lucknow, India. Prof. (Dr.) Vishnu Narayan Mishra, Assistant Professor of Mathematics, Sardar Vallabhbhai National Institute of Technology, Ichchhanath Mahadev Road, Surat, Surat-395007, Gujarat, India. Dr. Jia Hu, Member Research Staff, Philips Research North America, New York Area, NY. Prof. Shashikant Shantilal Patil SVKM , MPSTME Shirpur Campus, NMIMS University Vile Parle Mumbai, India. Prof. (Dr.) Bindhya Chal Yadav, Assistant Professor in Botany, Govt. Post Graduate College, Fatehabad, Agra, Uttar Pradesh, India. REVIEW BOARD Prof. (Dr.) Kimberly A. Freeman, Professor & Director of Undergraduate Programs, Stetson School of Business and Economics, Mercer University, Macon, Georgia, United States. Prof. (Dr.) Klaus G. Troitzsch, Professor, Institute for IS Research, University of Koblenz-Landau, Germany. Prof. (Dr.) T. Anthony Choi, Professor, Department of Electrical & Computer Engineering, Mercer University, Macon, Georgia, United States. Prof. (Dr.) Yen-Chun Lin, Professor and Chair, Dept. of Computer Science and Information Engineering, Chang Jung Christian University, Kway Jen, Tainan, Taiwan. Prof. (Dr.) Jen-Wei Hsieh, Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taiwan. Prof. (Dr.) Jose C. Martinez, Dept. Physical Chemistry, Faculty of Sciences, University of Granada, Spain. Prof. (Dr.) Joel Saltz, Emory University, Atlanta, Georgia, United States. Prof. (Dr.) Panayiotis Vafeas, Department of Engineering Sciences, University of Patras, Greece. Prof. (Dr.) Soib Taib, School of Electrical & Electronics Engineering, University Science Malaysia, Malaysia. Prof. (Dr.) Sim Kwan Hua, School of Engineering, Computing and Science, Swinburne University of Technology, Sarawak, Malaysia. Prof. (Dr.) Jose Francisco Vicent Frances, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Rafael Ignacio Alvarez Sanchez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Francisco Miguel Martinez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Antonio Zamora Gomez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Leandro Tortosa, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain. Prof. (Dr.) Samir Ananou, Department of Microbiology, Universidad de Granada, Granada, Spain. Dr. Miguel Angel Bautista, Department de Matematica Aplicada y Analisis, Facultad de Matematicas, Universidad de Barcelona, Spain. Prof. (Dr.) Prof. Adam Baharum, School of Mathematical Sciences, University of Universiti Sains, Malaysia, Malaysia. Prof. (Dr.) Huiyun Liu, Department of Electronic & Electrical Engineering, University College London, Torrington Place, London.

                                

Dr. Cristiano De Magalhaes Barros, Governo do Estado de Minas Gerais, Brazil. Prof. (Dr.) Pravin G. Ingole, Senior Researcher, Greenhouse Gas Research Center, Korea Institute of Energy Research (KIER), 152 Gajeong-ro, Yuseong-gu, Daejeon 305-343, KOREA. Prof. (Dr.) Dilum Bandara, Dept. Computer Science & Engineering, University of Moratuwa, Sri Lanka. Prof. (Dr.) Faudziah Ahmad, School of Computing, UUM College of Arts and Sciences, University Utara Malaysia, 06010 UUM Sintok, Kedah Darulaman. Prof. (Dr.) G. Manoj Someswar, Principal, Dept. of CSE at Anwar-ul-uloom College of Engineering & Technology, Yennepally, Vikarabad, RR District., A.P., India. Prof. (Dr.) Abdelghni Lakehal, Applied Mathematics, Rue 10 no 6 cite des fonctionnaires dokkarat 30010 Fes Marocco. Dr. Kamal Kulshreshtha, Associate Professor & Head, Deptt. of Computer Sc. & Applications, Modi Institute of Management & Technology, Kota-324 009, Rajasthan, India. Prof. (Dr.) Anukrati Sharma, Associate Professor, Faculty of Commerce and Management, University of Kota, Kota, Rajasthan, India. Prof. (Dr.) S. Natarajan, Department of Electronics and Communication Engineering, SSM College of Engineering, NH 47, Salem Main Road, Komarapalayam, Namakkal District, Tamilnadu 638183, India. Prof. (Dr.) J. Sadhik Basha, Department of Mechanical Engineering, King Khalid University, Abha, Kingdom of Saudi Arabia. Prof. (Dr.) G. SAVITHRI, Department of Sericulture, S.P. Mahila Visvavidyalayam, Tirupati517502, Andhra Pradesh, India. Prof. (Dr.) Shweta jain, Tolani College of Commerce, Andheri, Mumbai. 400001, India. Prof. (Dr.) Abdullah M. Abdul-Jabbar, Department of Mathematics, College of Science, University of Salahaddin-Erbil, Kurdistan Region, Iraq. Prof. (Dr.) ( Mrs.) P.Sujathamma, Department of Sericulture, S.P.Mahila Visvavidyalayam, Tirupati-517502, India. Prof. (Dr.) Bimla Dhanda, Professor & Head, Department of Human Development and Family Studies, College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India. Prof. (Dr.) Manjulatha, Dept of Biochemistry,School of Life Sciences,University of Hyderabad,Gachibowli, Hyderabad, India. Prof. (Dr.) Upasani Dhananjay Eknath Advisor & Chief Coordinator, ALUMNI Association, Sinhgad Institute of Technology & Science, Narhe, Pune -411 041, India. Prof. (Dr.) Sudhindra Bhat, Professor & Finance Area Chair, School of Business, Alliance University Bangalore-562106, India. Prof. Prasenjit Chatterjee , Dept. of Mechanical Engineering, MCKV Institute of Engineering West Bengal, India. Prof. Rajesh Murukesan, Deptt. of Automobile Engineering, Rajalakshmi Engineering college, Chennai, India. Prof. (Dr.) Parmil Kumar, Department of Statistics, University of Jammu, Jammu, India Prof. (Dr.) M.N. Shesha Prakash, Vice Principal, Professor & Head of Civil Engineering, Vidya Vikas Institute of Engineering and Technology, Alanahally, Mysore-570 028 Prof. (Dr.) Piyush Singhal, Mechanical Engineering Deptt., GLA University, India. Prof. M. Mahbubur Rahman, School of Engineering & Information Technology, Murdoch University, Perth Western Australia 6150, Australia. Prof. Nawaraj Chaulagain, Department of Religion, Illinois Wesleyan University, Bloomington, IL. Prof. Hassan Jafari, Faculty of Maritime Economics & Management, Khoramshahr University of Marine Science and Technology, khoramshahr, Khuzestan province, Iran Prof. (Dr.) Kantipudi MVV Prasad , Dept of EC, School of Engg., R.K.University, Kast urbhadham, Tramba, Rajkot-360020, India. Prof. (Mrs.) P.Sujathamma, Department of Sericulture, S.P.Mahila Visvavidyalayam, ( Women's University), Tirupati-517502, India. Prof. (Dr.) M A Rizvi, Dept. of Computer Engineering and Applications, National Institute of Technical Teachers' Training and Research, Bhopal M.P. India. Prof. (Dr.) Mohsen Shafiei Nikabadi, Faculty of Economics and Management, Industrial Management Department, Semnan University, Semnan, Iran. Prof. P.R.SivaSankar, Head, Dept. of Commerce, Vikrama Simhapuri University Post Graduate Centre, KAVALI - 524201, A.P. India. Prof. (Dr.) Bhawna Dubey, Institute of Environmental Science( AIES), Amity University, Noida, India. Prof. Manoj Chouhan, Deptt. of Information Technology, SVITS Indore, India.

                               

Prof. Yupal S Shukla, V M Patel College of Management Studies, Ganpat University, KhervaMehsana. India. Prof. (Dr.) Amit Kohli, Head of the Department, Department of Mechanical Engineering, D.A.V.Institute of Engg. and Technology, Kabir Nagar, Jalandhar,Punjab (India). Prof. (Dr.) Kumar Irayya Maddani, and Head of the Department of Physics in SDM College of Engineering and Technology, Dhavalagiri, Dharwad, State: Karnataka (INDIA). Prof. (Dr.) Shafi Phaniband, SDM College of Engineering and Technology, Dharwad, INDIA. Prof. M H Annaiah, Head, Department of Automobile Engineering, Acharya Institute of Technology, Soladevana Halli, Bangalore -560107, India. Prof. (Dr.) Prof. R. R. Patil, Director School Of Earth Science, Solapur University, Solapur Prof. (Dr.) Manoj Khandelwal, Dept. of Mining Engg, College of Technology & Engineering, Maharana Pratap University of Agriculture & Technology, Udaipur, 313 001 (Rajasthan), India Prof. (Dr.) Kishor Chandra Satpathy, Librarian, National Institute of Technology, Silchar-788010, Assam, India Prof. (Dr.) Juhana Jaafar, Gas Engineering Department, Faculty of Petroleum and Renewable Energy Engineering (FPREE), Universiti Teknologi Malaysia-81310 UTM Johor Bahru, Johor. Prof. (Dr.) Rita Khare, Assistant Professor in chemistry, Govt. Women’s College, Gardanibagh, Patna, Bihar. Prof. (Dr.) Raviraj Kusanur, Dept of Chemistry, R V College of Engineering, Bangalore-59, India. Prof. (Dr.) Hameem Shanavas .I, M.V.J College of Engineering, Bangalore Prof. (Dr.) Sanjay Kumar, JKL University, Ajmer Road, Jaipur Prof. (Dr.) Pushp Lata Faculty of English and Communication, Department of Humanities and Languages, Nucleus Member, Publications and Media Relations Unit Editor, BITScan, BITS, PilaniIndia. Prof. Arun Agarwal, Faculty of ECE Dept., ITER College, Siksha 'O' Anusandhan University Bhubaneswar, Odisha, India Prof. (Dr.) Pratima Tripathi, Department of Biosciences, SSSIHL, Anantapur Campus Anantapur515001 (A.P.) India. Prof. (Dr.) Sudip Das, Department of Biotechnology, Haldia Institute of Technology, I.C.A.R.E. Complex, H.I.T. Campus, P.O. Hit, Haldia; Dist: Puba Medinipur, West Bengal, India. Prof. (Dr.) Bimla Dhanda, Professor & Head, Department of Human Development and Family Studies College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India. Prof. (Dr.) R.K.Tiwari, Professor, S.O.S. in Physics, Jiwaji University, Gwalior, M.P.-474011. Prof. (Dr.) Deepak Paliwal, Faculty of Sociology, Uttarakhand Open University, Haldwani-Nainital Prof. (Dr.) Dr. Anil K Dwivedi, Faculty of Pollution & Environmental Assay Research Laboratory (PEARL), Department of Botany,DDU Gorakhpur University,Gorakhpur-273009,India. Prof. R. Ravikumar, Department of Agricultural and Rural Management, TamilNadu Agricultural University,Coimbatore-641003,TamilNadu,India. Prof. (Dr.) R.Raman, Professor of Agronomy, Faculty of Agriculture, Annamalai university, Annamalai Nagar 608 002Tamil Nadu, India. Prof. (Dr.) Ahmed Khalafallah, Coordinator of the CM Degree Program, Department of Architectural and Manufacturing Sciences, Ogden College of Sciences and Engineering Western Kentucky University 1906 College Heights Blvd Bowling Green, KY 42103-1066. Prof. (Dr.) Asmita Das , Delhi Technological University (Formerly Delhi College of Engineering), Shahbad, Daulatpur, Delhi 110042, India. Prof. (Dr.)Aniruddha Bhattacharjya, Assistant Professor (Senior Grade), CSE Department, Amrita School of Engineering , Amrita Vishwa VidyaPeetham (University), Kasavanahalli, Carmelaram P.O., Bangalore 560035, Karnataka, India. Prof. (Dr.) S. Rama Krishna Pisipaty, Prof & Geoarchaeologist, Head of the Department of Sanskrit & Indian Culture, SCSVMV University, Enathur, Kanchipuram 631561, India Prof. (Dr.) Shubhasheesh Bhattacharya, Professor & HOD(HR), Symbiosis Institute of International Business (SIIB), Hinjewadi, Phase-I, Pune- 411 057, India. Prof. (Dr.) Vijay Kothari, Institute of Science, Nirma University, S-G Highway, Ahmedabad 382481, India. Prof. (Dr.) Raja Sekhar Mamillapalli, Department of Civil Engineering at Sir Padampat Singhania University, Udaipur, India. Prof. (Dr.) B. M. Kunar, Department of Mining Engineering, Indian School of Mines, Dhanbad 826004, Jharkhand, India. Prof. (Dr.) Prabir Sarkar, Assistant Professor, School of Mechanical, Materials and Energy Engineering, Room 307, Academic Block, Indian Institute of Technology, Ropar, Nangal Road, Rupnagar 140001, Punjab, India.

                             

Prof. (Dr.) K.Srinivasmoorthy, Associate Professor, Department of Earth Sciences, School of Physical,Chemical and Applied Sciences, Pondicherry university, R.Venkataraman Nagar, Kalapet, Puducherry 605014, India. Prof. (Dr.) Bhawna Dubey, Institute of Environmental Science (AIES), Amity University, Noida, India. Prof. (Dr.) P. Bhanu Prasad, Vision Specialist, Matrix vision GmbH, Germany, Consultant, TIFACCORE for Machine Vision, Advisor, Kelenn Technology, France Advisor, Shubham Automation & Services, Ahmedabad, and Professor of C.S.E, Rajalakshmi Engineering College, India. Prof. (Dr.)P.Raviraj, Professor & Head, Dept. of CSE, Kalaignar Karunanidhi, Institute of Technology, Coimbatore 641402,Tamilnadu,India. Prof. (Dr.) Damodar Reddy Edla, Department of Computer Science & Engineering, Indian School of Mines, Dhanbad, Jharkhand 826004, India. Prof. (Dr.) T.C. Manjunath, Principal in HKBK College of Engg., Bangalore, Karnataka, India. Prof. (Dr.) Pankaj Bhambri, I.T. Deptt., Guru Nanak Dev Engineering College, Ludhiana 141006, Punjab, India. Prof. Shashikant Shantilal Patil SVKM , MPSTME Shirpur Campus, NMIMS University Vile Parle Mumbai, India. Prof. (Dr.) Shambhu Nath Choudhary, Department of Physics, T.M. Bhagalpur University, Bhagalpur 81200, Bihar, India. Prof. (Dr.) Venkateshwarlu Sonnati, Professor & Head of EEED, Department of EEE, Sreenidhi Institute of Science & Technology, Ghatkesar, Hyderabad, Andhra Pradesh, India. Prof. (Dr.) Saurabh Dalela, Department of Pure & Applied Physics, University of Kota, KOTA 324010, Rajasthan, India. Prof. S. Arman Hashemi Monfared, Department of Civil Eng, University of Sistan & Baluchestan, Daneshgah St.,Zahedan, IRAN, P.C. 98155-987 Prof. (Dr.) R.S.Chanda, Dept. of Jute & Fibre Tech., University of Calcutta, Kolkata 700019, West Bengal, India. Prof. V.S.VAKULA, Department of Electrical and Electronics Engineering, JNTUK, University College of Eng.,Vizianagaram5 35003, Andhra Pradesh, India. Prof. (Dr.) Nehal Gitesh Chitaliya, Sardar Vallabhbhai Patel Institute of Technology, Vasad 388 306, Gujarat, India. Prof. (Dr.) D.R. Prajapati, Department of Mechanical Engineering, PEC University of Technology,Chandigarh 160012, India. Dr. A. SENTHIL KUMAR, Postdoctoral Researcher, Centre for Energy and Electrical Power, Electrical Engineering Department, Faculty of Engineering and the Built Environment, Tshwane University of Technology, Pretoria 0001, South Africa. Prof. (Dr.)Vijay Harishchandra Mankar, Department of Electronics & Telecommunication Engineering, Govt. Polytechnic, Mangalwari Bazar, Besa Road, Nagpur- 440027, India. Prof. Varun.G.Menon, Department Of C.S.E, S.C.M.S School of Engineering, Karukutty,Ernakulam, Kerala 683544, India. Prof. (Dr.) U C Srivastava, Department of Physics, Amity Institute of Applied Sciences, Amity University, Noida, U.P-203301.India. Prof. (Dr.) Surendra Yadav, Professor and Head (Computer Science & Engineering Department), Maharashi Arvind College of Engineering and Research Centre (MACERC), Jaipur, Rajasthan, India. Prof. (Dr.) Sunil Kumar, H.O.D. Applied Sciences & Humanities Dehradun Institute of Technology, (D.I.T. School of Engineering), 48 A K.P-3 Gr. Noida (U.P.) 201308 Prof. Naveen Jain, Dept. of Electrical Engineering, College of Technology and Engineering, Udaipur-313 001, India. Prof. Veera Jyothi.B, CBIT, Hyderabad, Andhra Pradesh, India. Prof. Aritra Ghosh, Global Institute of Management and Technology, Krishnagar, Nadia, W.B. India Prof. Anuj K. Gupta, Head, Dept. of Computer Science & Engineering, RIMT Group of Institutions, Sirhind Mandi Gobindgarh, Punajb, India. Prof. (Dr.) Varala Ravi, Head, Department of Chemistry, IIIT Basar Campus, Rajiv Gandhi University of Knowledge Technologies, Mudhole, Adilabad, Andhra Pradesh- 504 107, India Prof. (Dr.) Ravikumar C Baratakke, faculty of Biology,Govt. College, Saundatti - 591 126, India. Prof. (Dr.) NALIN BHARTI, School of Humanities and Social Science, Indian Institute of Technology Patna, India. Prof. (Dr.) Shivanand S.Gornale , Head, Department of Studies in Computer Science, Government College (Autonomous), Mandya, Mandya-571 401-Karanataka, India.

                               

Prof. (Dr.) Naveen.P.Badiger, Dept.Of Chemistry, S.D.M.College of Engg. & Technology, Dharwad-580002, Karnataka State, India. Prof. (Dr.) Bimla Dhanda, Professor & Head, Department of Human Development and Family Studies, College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India. Prof. (Dr.) Tauqeer Ahmad Usmani, Faculty of IT, Salalah College of Technology, Salalah, Sultanate of Oman. Prof. (Dr.) Naresh Kr. Vats, Chairman, Department of Law, BGC Trust University Bangladesh Prof. (Dr.) Papita Das (Saha), Department of Environmental Science, University of Calcutta, Kolkata, India. Prof. (Dr.) Rekha Govindan , Dept of Biotechnology, Aarupadai Veedu Institute of technology , Vinayaka Missions University , Paiyanoor , Kanchipuram Dt, Tamilnadu , India. Prof. (Dr.) Lawrence Abraham Gojeh, Department of Information Science, Jimma University, P.o.Box 378, Jimma, Ethiopia. Prof. (Dr.) M.N. Kalasad, Department of Physics, SDM College of Engineering & Technology, Dharwad, Karnataka, India. Prof. Rab Nawaz Lodhi, Department of Management Sciences, COMSATS Institute of Information Technology Sahiwal. Prof. (Dr.) Masoud Hajarian, Department of Mathematics, Faculty of Mathematical Sciences, Shahid Beheshti University, General Campus, Evin, Tehran 19839,Iran Prof. (Dr.) Chandra Kala Singh, Associate professor, Department of Human Development and Family Studies, College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India Prof. (Dr.) J.Babu, Professor & Dean of research, St.Joseph's College of Engineering & Technology, Choondacherry, Palai,Kerala. Prof. (Dr.) Pradip Kumar Roy, Department of Applied Mechanics, Birla Institute of Technology (BIT) Mesra, Ranchi- 835215, Jharkhand, India. Prof. (Dr.) P. Sanjeevi kumar, School of Electrical Engineering (SELECT), Vandalur Kelambakkam Road, VIT University, Chennai, India. Prof. (Dr.) Debasis Patnaik, BITS-Pilani, Goa Campus, India. Prof. (Dr.) SANDEEP BANSAL, Associate Professor, Department of Commerce, I.G.N. College, Haryana, India. Dr. Radhakrishnan S V S, Department of Pharmacognosy, Faser Hall, The University of Mississippi Oxford, MS- 38655, USA. Prof. (Dr.) Megha Mittal, Faculty of Chemistry, Manav Rachna College of Engineering, Faridabad (HR), 121001, India. Prof. (Dr.) Mihaela Simionescu (BRATU), BUCHAREST, District no. 6, Romania, member of the Romanian Society of Econometrics, Romanian Regional Science Association and General Association of Economists from Romania Prof. (Dr.) Atmani Hassan, Director Regional of Organization Entraide Nationale Prof. (Dr.) Deepshikha Gupta, Dept. of Chemistry, Amity Institute of Applied Sciences,Amity University, Sec.125, Noida, India. Prof. (Dr.) Muhammad Kamruzzaman, Deaprtment of Infectious Diseases, The University of Sydney, Westmead Hospital, Westmead, NSW-2145. Prof. (Dr.) Meghshyam K. Patil , Assistant Professor & Head, Department of Chemistry,Dr. Babasaheb Ambedkar Marathwada University,Sub-Campus, Osmanabad- 413 501, Maharashtra, India. Prof. (Dr.) Ashok Kr. Dargar, Department of Mechanical Engineering, School of Engineering, Sir Padampat Singhania University, Udaipur (Raj.) Prof. (Dr.) Sudarson Jena, Dept. of Information Technology, GITAM University, Hyderabad, India Prof. (Dr.) Jai Prakash Jaiswal, Department of Mathematics, Maulana Azad National Institute of Technology Bhopal, India. Prof. (Dr.) S.Amutha, Dept. of Educational Technology, Bharathidasan University, Tiruchirappalli620 023, Tamil Nadu, India. Prof. (Dr.) R. HEMA KRISHNA, Environmental chemistry, University of Toronto, Canada. Prof. (Dr.) B.Swaminathan, Dept. of Agrl.Economics, Tamil Nadu Agricultural University, India. Prof. (Dr.) K. Ramesh, Department of Chemistry, C.B.I.T, Gandipet, Hyderabad-500075. India. Prof. (Dr.) Sunil Kumar, H.O.D. Applied Sciences &Humanities, JIMS Technical campus,(I.P. University,New Delhi), 48/4 ,K.P.-3,Gr.Noida (U.P.) Prof. (Dr.) G.V.S.R.Anjaneyulu, CHAIRMAN - P.G. BOS in Statistics & Deputy Coordinator UGC DRS-I Project, Executive Member ISPS-2013, Department of Statistics, Acharya Nagarjuna University, Nagarjuna Nagar-522510, Guntur, Andhra Pradesh, India.

                                

Prof. (Dr.) Sribas Goswami, Department of Sociology, Serampore College, Serampore 712201, West Bengal, India. Prof. (Dr.) Sunanda Sharma, Department of Veterinary Obstetrics Y Gynecology, College of Veterinary & Animal Science,Rajasthan University of Veterinary & Animal Sciences,Bikaner334001, India. Prof. (Dr.) S.K. Tiwari, Department of Zoology, D.D.U. Gorakhpur University, Gorakhpur-273009 U.P., India. Prof. (Dr.) Praveena Kuruva, Materials Research Centre, Indian Institute of Science, Bangalore560012, INDIA Prof. (Dr.) Rajesh Kumar, Department Of Applied Physics, Bhilai Institute Of Technology, Durg (C.G.) 491001, India. Dr. K.C.Sivabalan, Field Enumerator and Data Analyst, Asian Vegetable Research Centre, The World Vegetable Centre, Taiwan. Prof. (Dr.) Amit Kumar Mishra, Department of Environmntal Science and Energy Research, Weizmann Institute of Science, Rehovot, Israel. Prof. (Dr.) Manisha N. Paliwal, Sinhgad Institute of Management, Vadgaon (Bk), Pune, India. Prof. (Dr.) M. S. HIREMATH, Principal, K.L.ESOCIETY’s SCHOOL, ATHANI Prof. Manoj Dhawan, Department of Information Technology, Shri Vaishnav Institute of Technology & Science, Indore, (M. P.), India. Prof. (Dr.) V.R.Naik, Professor & Head of Department, Mechancal Engineering, Textile & Engineering Institute, Ichalkaranji (Dist. Kolhapur), Maharashatra, India. Prof. (Dr.) Jyotindra C. Prajapati,Head, Department of Mathematical Sciences, Faculty of Applied Sciences, Charotar University of Science and Technology, Changa Anand -388421, Gujarat, India Prof. (Dr.) Sarbjit Singh, Head, Department of Industrial & Production Engineering, Dr BR Ambedkar National Institute of Technology,Jalandhar,Punjab, India. Prof. (Dr.) Professor Braja Gopal Bag, Department of Chemistry and Chemical Technology , Vidyasagar University, West Midnapore Prof. (Dr.) Ashok Kumar Chandra, Department of Management, Bhilai Institute of Technology, Bhilai House, Durg (C.G.) Prof. (Dr.) Amit Kumar, Assistant Professor, School of Chemistry, Shoolini University, Solan, Himachal Pradesh, India Prof. (Dr.) L. Suresh Kumar, Mechanical Department, Chaitanya Bharathi Institute of Technology, Hyderabad, India. Scientist Sheeraz Saleem Bhat, Lac Production Division, Indian Institute of Natural Resins and Gums, Namkum, Ranchi, Jharkhand, India. Prof. C.Divya , Centre for Information Technology and Engineering, Manonmaniam Sundaranar University, Tirunelveli - 627012, Tamilnadu , India. Prof. T.D.Subash, Infant Jesus College Of Engineering and Technology, Thoothukudi Tamilnadu, India. Prof. (Dr.) Vinay Nassa, Prof. E.C.E Deptt., Dronacharya.Engg. College, Gurgaon India. Prof. Sunny Narayan, university of Roma Tre, Italy. Prof. (Dr.) Sanjoy Deb, Dept. of ECE, BIT Sathy, Sathyamangalam, Tamilnadu-638401, India. Prof. (Dr.) Reena Gupta, Institute of Pharmaceutical Research, GLA University, Mathura, India. Prof. (Dr.) P.R.SivaSankar, Head Dept. of Commerce, Vikrama Simhapuri University Post Graduate Centre, KAVALI - 524201, A.P., India. Prof. (Dr.) Mohsen Shafiei Nikabadi, Faculty of Economics and Management, Industrial Management Department, Semnan University, Semnan, Iran. Prof. (Dr.) Praveen Kumar Rai, Department of Geography, Faculty of Science, Banaras Hindu University, Varanasi-221005, U.P. India. Prof. (Dr.) Christine Jeyaseelan, Dept of Chemistry, Amity Institute of Applied Sciences, Amity University, Noida, India. Prof. (Dr.) M A Rizvi, Dept. of Computer Engineering and Applications , National Institute of Technical Teachers' Training and Research, Bhopal M.P. India. Prof. (Dr.) K.V.N.R.Sai Krishna, H O D in Computer Science, S.V.R.M.College,(Autonomous), Nagaram, Guntur(DT), Andhra Pradesh, India. Prof. (Dr.) Ashok Kr. Dargar, Department of Mechanical Engineering, School of Engineering, Sir Padampat Singhania University, Udaipur (Raj.) Prof. (Dr.) Asim Kumar Sen, Principal , ST.Francis Institute of Technology (Engineering College) under University of Mumbai , MT. Poinsur, S.V.P Road, Borivali (W), Mumbai-400103, India. Prof. (Dr.) Rahmathulla Noufal.E, Civil Engineering Department, Govt.Engg.College-Kozhikode

                                

Prof. (Dr.) N.Rajesh, Department of Agronomy, TamilNadu Agricultural University -Coimbatore, Tamil Nadu, India. Prof. (Dr.) Har Mohan Rai , Professor, Electronics and Communication Engineering, N.I.T. Kurukshetra 136131,India Prof. (Dr.) Eng. Sutasn Thipprakmas from King Mongkut, University of Technology Thonburi, Thailand. Prof. (Dr.) Kantipudi MVV Prasad, EC Department, RK University, Rajkot. Prof. (Dr.) Jitendra Gupta,Faculty of Pharmaceutics, Institute of Pharmaceutical Research, GLA University, Mathura. Prof. (Dr.) Swapnali Borah, HOD, Dept of Family Resource Management, College of Home Science, Central Agricultural University, Tura, Meghalaya, India. Prof. (Dr.) N.Nazar Khan, Professor in Chemistry, BTK Institute of Technology, Dwarahat-263653 (Almora), Uttarakhand-India. Prof. (Dr.) Rajiv Sharma, Department of Ocean Engineering, Indian Institute of Technology Madras, Chennai (TN) - 600 036,India. Prof. (Dr.) Aparna Sarkar,PH.D. Physiology, AIPT,Amity University , F 1 Block, LGF, Sector125,Noida-201303, UP ,India. Prof. (Dr.) Manpreet Singh, Professor and Head, Department of Computer Engineering, Maharishi Markandeshwar University, Mullana, Haryana, India. Prof. (Dr.) Sukumar Senthilkumar, Senior Researcher Advanced Education Center of Jeonbuk for Electronics and Information Technology, Chon Buk National University, Chon Buk, 561-756, SOUTH KOREA. . Prof. (Dr.) Hari Singh Dhillon, Assistant Professor, Department of Electronics and Communication Engineering, DAV Institute of Engineering and Technology, Jalandhar (Punjab), INDIA. . Prof. (Dr.) Poonkuzhali, G., Department of Computer Science and Engineering, Rajalakshmi Engineering College, Chennai, INDIA. . Prof. (Dr.) Bharath K N, Assistant Professor, Dept. of Mechanical Engineering, GM Institute of Technology, PB Road, Davangere 577006, Karnataka, INDIA. . Prof. (Dr.) F.Alipanahi, Assistant Professor, Islamic Azad University,Zanjan Branch, Atemadeyeh, Moalem Street, Zanjan IRAN Prof. Yogesh Rathore, Assistant Professor, Dept. of Computer Science & Engineering, RITEE, Raipur, India Prof. (Dr.) Ratneshwer, Department of Computer Science (MMV), Banaras Hindu University Varanasi-221005, India. Prof. Pramod Kumar Pandey, Assistant Professor, Department Electronics & Instrumentation Engineering, ITM University, Gwalior, M.P., India Prof. (Dr.)Sudarson Jena, Associate Professor, Dept.of IT, GITAM University, Hyderabad, India Prof. (Dr.) Binod Kumar,PhD(CS), M.Phil(CS),MIEEE,MIAENG, Dean & Professor( MCA), Jayawant Technical Campus(JSPM's), Pune, India Prof. (Dr.) Mohan Singh Mehata, (JSPS fellow), Assistant Professor, Department of Applied Physics, Delhi Technological University, Delhi Prof. Ajay Kumar Agarwal, Asstt. Prof., Deptt. of Mech. Engg., Royal Institute of Management & Technology, Sonipat (Haryana) Prof. (Dr.) Siddharth Sharma, University School of Management, Kurukshetra University, Kurukshetra, India. Prof. (Dr.) Satish Chandra Dixit, Department of Chemistry, D.B.S.College ,Govind Nagar,Kanpur208006, India Prof. (Dr.) Ajay Solkhe, Department of Management, Kurukshetra University, Kurukshetra, India. Prof. (Dr.) Neeraj Sharma, Asst. Prof. Dept. of Chemistry, GLA University, Mathura Prof. (Dr.) Basant Lal, Department of Chemistry, G.L.A. University, Mathura Prof. (Dr.) T Venkat Narayana Rao, C.S.E,Guru Nanak Engineering College, Hyderabad, Andhra Pradesh, India Prof. (Dr.) Rajanarender Reddy Pingili, S.R. International Institute of Technology, Hyderabad, Andhra Pradesh, India Prof. (Dr.) V.S.Vairale, Department of Computer Engineering, All India Shri Shivaji Memorial Society College of Engineering, Kennedy Road, Pune-411 001, Maharashtra, India Prof. (Dr.) Vasavi Bande, Department of Computer Science & Engineering, Netaji Institute of Engineering and Technology, Hyderabad, Andhra Pradesh, India Prof. (Dr.) Hardeep Anand, Department of Chemistry, Kurukshetra University Kurukshetra, Haryana, India. Prof. Aasheesh shukla, Asst Professor, Dept. of EC, GLA University, Mathura, India.

                                                

Prof. S.P.Anandaraj., CSE Dept, SREC, Warangal, India. Satya Rishi Takyar , Senior ISO Consultant, New Delhi, India. Prof. Anuj K. Gupta, Head, Dept. of Computer Science & Engineering, RIMT Group of Institutions, Mandi Gobindgarh, Punjab, India. Prof. (Dr.) Harish Kumar, Department of Sports Science, Punjabi University, Patiala, Punjab, India. Prof. (Dr.) Mohammed Ali Hussain, Professor, Dept. of Electronics and Computer Engineering, KL University, Green Fields, Vaddeswaram, Andhra Pradesh, India. Prof. (Dr.) Manish Gupta, Department of Mechanical Engineering, GJU, Haryana, India. Prof. Mridul Chawla, Department of Elect. and Comm. Engineering, Deenbandhu Chhotu Ram University of Science & Technology, Murthal, Haryana, India. Prof. Seema Chawla, Department of Bio-medical Engineering, Deenbandhu Chhotu Ram University of Science & Technology, Murthal, Haryana, India. Prof. (Dr.) Atul M. Gosai, Department of Computer Science, Saurashtra University, Rajkot, Gujarat, India. Prof. (Dr.) Ajit Kr. Bansal, Department of Management, Shoolini University, H.P., India. Prof. (Dr.) Sunil Vasistha, Mody Institute of Tecnology and Science, Sikar, Rajasthan, India. Prof. Vivekta Singh, GNIT Girls Institute of Technology, Greater Noida, India. Prof. Ajay Loura, Assistant Professor at Thapar University, Patiala, India. Prof. Sushil Sharma, Department of Computer Science and Applications, Govt. P. G. College, Ambala Cantt., Haryana, India. Prof. Sube Singh, Assistant Professor, Department of Computer Engineering, Govt. Polytechnic, Narnaul, Haryana, India. Prof. Himanshu Arora, Delhi Institute of Technology and Management, New Delhi, India. Dr. Sabina Amporful, Bibb Family Practice Association, Macon, Georgia, USA. Dr. Pawan K. Monga, Jindal Institute of Medical Sciences, Hisar, Haryana, India. Dr. Sam Ampoful, Bibb Family Practice Association, Macon, Georgia, USA. Dr. Nagender Sangra, Director of Sangra Technologies, Chandigarh, India. Vipin Gujral, CPA, New Jersey, USA. Sarfo Baffour, University of Ghana, Ghana. Monique Vincon, Hype Softwaretechnik GmbH, Bonn, Germany. Natasha Sigmund, Atlanta, USA. Marta Trochimowicz, Rhein-Zeitung, Koblenz, Germany. Kamalesh Desai, Atlanta, USA. Vijay Attri, Software Developer Google, San Jose, California, USA. Neeraj Khillan, Wipro Technologies, Boston, USA. Ruchir Sachdeva, Software Engineer at Infosys, Pune, Maharashtra, India. Anadi Charan, Senior Software Consultant at Capgemini, Mumbai, Maharashtra. Pawan Monga, Senior Product Manager, LG Electronics India Pvt. Ltd., New Delhi, India. Sunil Kumar, Senior Information Developer, Honeywell Technology Solutions, Inc., Bangalore, India. Bharat Gambhir, Technical Architect, Tata Consultancy Services (TCS), Noida, India. Vinay Chopra, Team Leader, Access Infotech Pvt Ltd. Chandigarh, India. Sumit Sharma, Team Lead, American Express, New Delhi, India. Vivek Gautam, Senior Software Engineer, Wipro, Noida, India. Anirudh Trehan, Nagarro Software Gurgaon, Haryana, India. Manjot Singh, Senior Software Engineer, HCL Technologies Delhi, India. Rajat Adlakha, Senior Software Engineer, Tech Mahindra Ltd, Mumbai, Maharashtra, India. Mohit Bhayana, Senior Software Engineer, Nagarro Software Pvt. Gurgaon, Haryana, India. Dheeraj Sardana, Tech. Head, Nagarro Software, Gurgaon, Haryana, India. Naresh Setia, Senior Software Engineer, Infogain, Noida, India. Raj Agarwal Megh, Idhasoft Limited, Pune, Maharashtra, India. Shrikant Bhardwaj, Senior Software Engineer, Mphasis an HP Company, Pune, Maharashtra, India. Vikas Chawla, Technical Lead, Xavient Software Solutions, Noida, India. Kapoor Singh, Sr. Executive at IBM, Gurgaon, Haryana, India. Ashwani Rohilla, Senior SAP Consultant at TCS, Mumbai, India. Anuj Chhabra, Sr. Software Engineer, McKinsey & Company, Faridabad, Haryana, India. Jaspreet Singh, Business Analyst at HCL Technologies, Gurgaon, Haryana, India.

TOPICS OF INTEREST Topics of interest include, but are not limited to, the following:  Social networks and intelligence  Social science simulation  Information retrieval systems  Technology management  Digital libraries for e-learning  Web-based learning, wikis and blogs  Operational research  Ontologies and meta-data standards  Engineering problems and emerging application  Agent based modeling and systems  Ubiquitous computing  Wired and wireless data communication networks  Mobile Ad Hoc, sensor and mesh networks  Natural language processing and expert systems  Monte Carlo methods and applications  Fuzzy logic and soft computing  Data mining and warehousing  Software and web engineering  Distributed AI systems and architectures  Neural networks and applications  Search and meta-heuristics  Bioinformatics and scientific computing  Genetic network modeling and inference  Knowledge and information management techniques  Aspect-oriented programming  Formal and visual specification languages  Informatics and statistics research  Quantum computing  Automata and formal languages  Computer graphics and image processing  Web 3D and applications  Grid computing and cloud computing  Algorithms design  Genetic algorithms  Compilers and interpreters  Computer architecture & VLSI  Advanced database systems  Digital signal and image processing  Distributed and parallel processing  Information retrieval systems  Technology management  Automation and mobile robots  Manufacturing technology  Electrical technology  Applied mathematics  Automatic control  Nuclear engineering  Computational physics  Computational chemistry

TABLE OF CONTENTS (December-2013 to February-2014, Issue 7, Volume 1, 2, 3 & 4) Issue 7, Volume 1 Paper Code

Paper Title

Page No.

IJETCAS 14-101

A Novel Model for Implementing Security over Mobile Ad-hoc Networks using Intuitionistic Fuzzy Function A. A. Salama, M.Abdelfattah, Y .M. Wazery

01-07

IJETCAS 14-102

Automatic Detection and Classification of Glioma Tumors using Statistical Features Ananda Resmi S., Tessamma Thomas

08-14

IJETCAS 14-103

Comparative studies of Cerium and Zirconium doped Barium Titanate S.N. Rahman, N. Khatun, S. Islam, N.A. Ahmed

15-19

IJETCAS 14-104

Structural and Electrical Characterization of Ni-Zn Ferrites Md. Shahjahan, N. A. Ahmed, S. N. Rahman, S. Islam, N. Khatun

20-25

IJETCAS 14-106

Comparative Study of Web Page Ranking Algorithms Atul Kumar Srivastava, Mitali Srivastava, Rakhi Garg, P. K. Mishra

26-32

IJETCAS 14-107

An Experimental Investigation of the Hydraulic and Durability Properties of Cement Treated Permeable Basecourses (CTPBs) Abdul A. Koroma, Victor S. Kamara

33-39

IJETCAS 14-108

Effects of Ni and Cd contaminated fish meat on chromosomal aberrations and sperm morphology of Swiss albino mice Mustafa S.Al-Attar and Rezan O. Rasheed

40-44

IJETCAS 14-109

Evaluation of the Uncertainty in Spectral Peak Location Case Study: Symmetrical Lines Prof. J. Dubrovkin

45-53

IJETCAS 14-111

A ROBUST APPROACH FOR OBJECT TRACKING BASED ON PARTICLE FILTER AND OPTIMIZED LIKELIHOOD Amr M. Nagy, Ali Ahmed and Hala H. Zayed

54-61

IJETCAS 14-113

Star-Mobius Cube: A New Interconnection Topology for Large Scale Parallel Processing Debasmita Pattanayak, Devashree Tripathy and C.R.Tripathy

62-68

IJETCAS 14-114

Diffraction Ring Technique and Nonlinear Optical Properties of 5-Aminoindazole Abdulameer Imran, Hussain A. Badran, Qusay Mohammed Ali Hassan

69-74

IJETCAS 14-115

Seismic Data Analysis in Odyssey Software Nazarov Yuri P., Poznyak Elena V., Filimonov Anton V.

75-77

IJETCAS 14-116

KINETICS AND MECHANISM OF ZnCl2 CATALYSED OXIDATION OF ORGANIC SULFOXIDES BY PERMANGANATE IN NON-AQUEOUS MEDIUM K.P.Srivastava & Sanjay Kumar Rai

78-85

IJETCAS 14-117

A STUDY ON IMAGE DENOISING FOR LUNG CT SCAN IMAGES S.Sivakumar and Dr.C.Chandrasekar

86-91

IJETCAS 14-119

A Novel Bit Error Rate Reduction Method for 3GPP-LTE-SCFDMA Using the Multiwavelet Transform Raad Farhood Chisab, Prof. (Dr.) C. K. Shukla

92-100

IJETCAS 14-121

A Novel Method for Image Segmentation Using Fuzzy Threshold Selection Janakiraman.S, J.Gowri

101-105

IJETCAS 14-122

A Hybrid Approach to normalize the light illumination in facial images using DCT and Gamma Transformation C.Arunkumar, T.Raghuram, M.N.Sekharan

106-112

IJETCAS 14-123

Review on Scuderi Split Cycle Engine Ashwini S. Gaikwad, Rajendra M.Shinde

113-117

Issue 7, Volume 2 Paper Code

Paper Title

Page No.

IJETCAS 14-124

Recognition of Facial Expressions With Respect to Navarasas in Bharathanatyam Styles Using Neural Network Prof. Dr. P. K. Srimani, Mr. Ramesh Hegde

118-121

IJETCAS 14-125

DESIGNING PARABOLIC PULSE WIDTH MODULATION FOR INDUCTION MOTOR DRIVE A.Elakya, G.S.Arun Kumar

122-126

IJETCAS 14-126

A Comparative Study of Association Rule Mining Algorithms on Grid and Cloud Platform Sudhakar Singh, Rakhi Garg, P. K. Mishra

127-134

IJETCAS 14-129

COMPARATIVE ANALYSIS OF APRIORI AND IMPROVED APRIORI ALGORITHM Prafulla Bafna , Pravin Metkewar and Angelina Gokhale

135-143

IJETCAS 14-131

Analysis of Brain tumor using neural network based PCA and clustering Prof.Dr.P.K.Srimani, Prof.Shanthi Mahesh, Dr.Neha Mangla

144-149

IJETCAS 14-132

Design and Simulation of Fuzzy Implication Function of Fuzzy System Using Two Stage CMOS Operational Amplifier Shruti Jain

150-155

IJETCAS 14-133

A Survey on FPGA implementation of 3D DWT using Lifting based Algorithm Hamsavahini R, Naveena.S,Meti Reddy Kshetra, Chaitra.R, Chaithra.K

156-161

IJETCAS 14-134

Review of Existing Security Protocols Techniques and their Performance Analysis in WLAN Latha P.H, Dr. Vasantha R

162-171

IJETCAS 14-135

A BUILT-IN REPAIR ANALYZER WITH SYNDROME STORING AND BIT-SWAPPING LFSR IN MEMORIES S.Jeevitha, R.Yamini

172-176

IJETCAS 14-136

Enhanced Local Search in Artificial Bee Colony Algorithm Sandeep Kumar, Dr. Vivek Kumar Sharma

177-184

IJETCAS 14-137

Design of Rule Scheduler for Trigger Rule Conflict in Active Object-Relational Database Systems Ms. S.Meenakshi, Dr. V. Thiagarasu

185-189

IJETCAS 14-138

STABILITY IN A DISCRETE NONLINEAR PREY - PREDATOR MODEL WITH FUNCTIONAL RESPONSE M.ReniSagaya Raj, A.George Maria Selvam and R.Dhineshbabu

190-197

IJETCAS 14-141

Modified Position Update in Spider Monkey Optimization Algorithm Sandeep Kumar, Dr. Vivek Kumar Sharma, Rajani Kumari

198-204

IJETCAS 14-142

Big Data Approach to Analytical Chemistry J. Dubrovkin

205-210

IJETCAS 14-143

The Crossed cube-Mesh: A New Fault-Tolerant Interconnection Network Topology for Parallel Systems Bhaskar Jyoti Das, Devashree Tripathy and Bibek Mishra

211-219

IJETCAS 14-144

MINIMIZATION OF DYNAMIC POWER USING PULSED LATCH FOR CLOCK TREE T.Angala parameshwari, T.Thamaraimanalan

220-224

IJETCAS 14-145

Particle Swarm Optimization Algorithm for Designing Optimal IIR Digital Filter Ranjit Kaur, Damanpreet Singh

225-230

IJETCAS 14-146

Real-time Compressive Tracking - A Study and Review Dr.S.Ravi, Salim Guria, P.Pasupathi, S.Muthukumar

231-236

IJETCAS 14-147

SELFISH CARRIER MONITORING IN WIFI USING DISTRIBUTED SNIFFERS U.SINTHUJA, R.SRIDEVI

237-241

Issue 7, Volume 3 Paper Code

Paper Title

Page No.

IJETCAS 14-148

Newtonian heating and mass transfer effects of chemical reactive fluid on natural convection flow past a vertical surface with radiation ARPITA JAIN

242-246

IJETCAS 14-149

BIANCHI TYPE V COSMOLOGICAL MODEL WITH DECAYING COSMOLOGICAL CONSTANT TERM R.K. Tiwari, Sudha Agrawal & D.K.Tiwari

247-251

IJETCAS 14-150

AN EFFICIENT GENE SELECTION FOR MICROARRAY DATA USING CLUSTERING ALGORITHM G. BASKAR, Dr.P.PONMUTHURAMALINGAM

252-254

IJETCAS 14-151

Test Suites Prioritization for Regression Testing using Genetic Algorithm G.N Purohit, A.M Sherry, Manish Saraswat

255-259

IJETCAS 14-152

EFFICIENT LOCALIZATION SCHEME FOR A LOW POWER CONTENT ADDRESSABLE MEMORY BASED ON XNOR CELL

260-264

G.Visnu Priya , S.Kokilamani, Ass.Prof. L.Jubair Ahmed IJETCAS 14-153

Implementation of Reverse Vending Machine Based on FPGA Sakshi Sharma, Ana Monga

265-271

IJETCAS 14-154

CBIR: Repossession of Analogous Pictures exhausting Median Vector Algorithm Dr.S.Janakiraman, Suthir.S

272-277

IJETCAS 14-156

Use of Genetic Algorithm in Web Information Retrieval Mr. Vikas Thada, Dr. Vivek Jaglan

278-281

IJETCAS 14-161

TO MEASURE THE EXTENT OF VARIATION IN MAJOR PULSE CROP D.P. Singh, M.S. Paikra, A.K. Thakur and Deo Shankar

282-284

IJETCAS 14-162

STUDY OF CONSCIOUSNESS IN TRADITIONAL LEARNING EXPLORE WITH CLASSIFICATION STRATEGY D. Magdalene Delighta Angeline, Dr. P. Ramasubramanian, I. Samuel Peter James

285-289

IJETCAS 14-164

Copy-Move Forgery Detection Utilizing Local Binary Patterns Ramandeep Kaur

290-294

IJETCAS 14-165

Study of Urbanization and its effect on Water resources in Yale Mallappa Shetty Kere watershed, Bangalore North taluk using Remote Sensing and GIS Chandan G.S and Renuka Prasad T.J

295-300

IJETCAS 14-166

SYNTHESIS AND ANTIMICROBIAL ACTIVITY OF METALLIC SOAPS OF LOWER CARBOXILIC ACIDS Binki Gangwar, Rajesh Dwivedi, Neeraj Sharma and Meera Sharma

301-305

IJETCAS 14-167

Review on Path Traversal for Web Navigation Mining Sagar More, Suraj Patil, Nihit Agrawal, Rohit Pachlor

306-309

IJETCAS 14-168

Molecular Interactions in Certain Binary Liquid Systems Involving Toluene at Temperature 313K Y. C. Morey and P. S. Agrawal

310-316

IJETCAS 14-169

Effect of antibiotics on varied density of septicemia causing bacteria of silkworm H. K. Ashok Kumar, S. Ramakrishna

317-320

IJETCAS 14-170

Performance Analysis and Architectural Overview of SAP HANA AYUSHI CHHABRA, SHWETANK SHARMA, SANJAY OJHA

321-326

IJETCAS 14-171

Comparison of Geranium essential Oil Production from different leaf sizes at constant Power through Solvent Free Microwave Extraction Method Ashutosh Mishra and Dr. S K Garg

327-331

IJETCAS 14-172

Cropping intensity, productivity, agricultural development and planning as influenced by integrated water resource management Anadi Gayen and A. Zaman

332-335

IJETCAS 14-173

Kinetics and mechanism of electron transfer to Sodiumhexachloroiridate (IV) from N, Nethylenethiourea in aqueous perchlorate media Sadhana Senapati, Smrutiprava Das, S.C. Dash, P. Mohanty, A.K. Pattnaik

336-341

Issue 7, Volume 4 Paper Code

Paper Title

Page No.

IJETCAS 14-174

A Novel and Simpler Approach for Ranking of Indoor Air Pollutants Rajesh Gopinath and Aftab Jahan Begum K.A.

342-344

IJETCAS 14-175

Designing An Intelligent Street Lighting system using a Zigbee Network Brinda.N, Thamaraimanalan.T, Nivethitha.T

345-350

IJETCAS 14-176

FILMED TRANSPORTATION INVESTIGATION FOR STRANGE AFFAIR RECOGNITION USING NUMEROUS ENTRY TRADITIONAL EXCAVATION Dr.S.Janakiraman, Suthir.S

351-358

IJETCAS 14-179

A Genetic Algorithm for K-Mean Clustering Varsha Singh, Prof A K Misra

359-364

IJETCAS 14-180

A Study on the Academic Performance of the Students by Applying K-Means Algorithm G. Narasinga Rao

365-366

IJETCAS 14-181

Semantic Web user personalization and search techniques based on time and weather condition R. Kousalya, Dr.V.Saravanan

367-371

IJETCAS 14-182

Molecular Orbital Studies on the isomers of mono aminophenol Linu Sam, Amrutha.R and P.Chandran

372-376

IJETCAS 14-184

Determining safety and Stability of Overhang Pressure Vessel using Finite Element Analysis Rohit Upadhyay, Prashant Bajaj, Vinay Patil

377-380

IJETCAS 14-185

Tamil Phonemes Recognition for Normal and Hearing Impaired: A Survey Janakiraman.S, Pavithra.S

381-387

IJETCAS 14-187

Analysis of Low Noise Amplifier in 1-100 GHz Range Somit Pandey

388-390

IJETCAS 14-190

VAIDYA-ANWESHANAM: One Stop Solution for All Medical Needs Shwetank Sharma, Kapil, Sanjay Ojha

391-394

IJETCAS 14-195

DESIGN OF EFFICIENT CLOCK GATED DET-FLIPFLOP FOR ASYNCHRONOUS DATA SAMPLING M.Keerthika, M.Ramya, S.Udhaya Kumar

395-399

IJETCAS 14-198

Simulation Study of Receive Diversity Techniques for MIMO Wireless Communication System Using MATLAB Prabhat Kumar, Manisha Manoj Chande

400-403

IJETCAS 14-203

Review on different types of Junctionless Transistors Porag Jyoti Ligira Gargi Khanna

404-408

IJETCAS 14-204

Applying data mining technique to analyze the causes of backlogs in University Rajiv Senapati, Dr. G.Sambasiva Rao

409-411

IJETCAS 14-213

Web Based Remote Lighting System Using ZigBee Technology and SoC Solution Alok Karnik, Ketan Khandagale, Pratik Jambhulkar

412-415

IJETCAS 14-219

Analysis of EEG signal for detection of Epilepsy Seizure Medha Jayant Patil and Bharati K. Khadse

416-420

IJETCAS 14-223

An Improved Feature Subset Selection Method Using Clusters S. Kowsalya, A. Nagajothi

421-426

IJETCAS 14-224

SVM-based Classifier System for Recognition of Cotton Leaf Diseases P.R. Rothe and Dr. R. V. Kshirsagar

427-432

IJETCAS 14-225

A Study on Techniques and Architecture of Sentiment Analysis Hina Jain, Yogesh Kumar Meena

433-437

IJETCAS 14-231

Investigation of Physical, Mechanical, and Morphological Properties of Acrylonitrile Butadiene Styrene on Unsaturated Polyester Resins D., Nouzari, O., Arjmand, M., Mouavi

438-443

IJETCAS 14-235

A SMART HOME MONITORING SYSTEM FOR ELDERLY PEOPLE B.Nandhini & R.Janani

444-447

IJETCAS

An Experimentation Work on Sn-ZnO using Sol-gel and characterization using XRD and AFM Patterns

14-236

Anand, Rajesh Malik,Rajesh Khanna

IJETCAS

Characterization of Deposition of ZnO using Sol-Gel approach

14-237

Suraksha, Rajesh Malik,Rajesh Khanna

IJETCAS

Solutal Convection in a Gravity Modulated Mushy Layer During the Solidification of Binary Alloys

14-238

Prof. Dr. P.K. Srimani1 and G. S. Mytra2

448-452

453-456

457-462

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A Novel Model for Implementing Security over Mobile Ad-hoc Networks using Intuitionistic Fuzzy Function 3

A. A. Salama1, Mohamed Abdelfattah2, Mohamed Eisa Math and Computer Science Department, Faculty of Science Port Said University, EGYPT 2 Information System Department, Faculty of Computers & Information, Benha University, EGYPT 3 Computer Science Department, Faculty of Science Port Said University, EGYPT 1

Abstract: Mobile adhoc network is a special kind of wireless networks. It is a collection of mobile nodes without having aid of establish infrastructure. In mobile adhoc network, it is much more vulnerable to attacks than a wired network due to its limited physical security, Securing temporal networks like Mobile Ad-hoc Networks (MANETs) has been given a great amount of attention recently, though the process of creating a perfectly secured scheme has not been accomplished yet. MANETs has some other features and characteristics those are together make it a difficult environment to be secured. The bandwidth of MANET is another challenge because it is unlikely to consume the bandwidth in security mechanisms rather than data traffic. This paper proposes a security scheme based on Public Key infrastructure (PKI) for distributing session keys between nodes. The length of those keys is decided using intuitionistic fuzzy logic manipulation. The proposed algorithm of Securitymodel is an adaptive intuitionistic fuzzy logic based algorithm that can adapt itself according to the dynamic conditions of mobile hosts. Finally the Experimental results shows that the using of intuitionistic fuzzy based security can enhance the security of (MANETs). Keywords: MANET; Security; wireless Communication; intuitionistic fuzzy; PKI; KNN I. Introduction Adhoc is a Latin word that means "for this or that only" AdHoc Networks, as its name indicates, are "intended to be" temporary. The idea is to completely remove any Base Station. Imagine a scenario in a relief operation in the event of timely communication is a very important factor, aid workers in the area are without the need of any existing infrastructure, just turn on the phone and start communicating with each other during movement and the execution of rescue operations [1]. A major challenge in the design of these networks is their vulnerability to security attacks. This article presents an overview of the security and ad hoc networks, and security threats applicable to ad hoc networks. It proposed a wide range of military and commercial applications for MANET. For example, a unit of soldiers that move in the battlefield cannot afford to install a base station every time you go to a new area. Similarly, the creation of a communication infrastructure for conference meeting informal and spontaneous between a small number of people that cannot be economically justified [2]. Even the robot-based networks in which multiple robots work at the same time to make the piles are extremely difficult for humans (the discovery of outer space and the extraction of minerals), smart homes and other important applications known to exist in applications vehicles Auto-routing. In addition, MANET can be the perfect tool for disaster recovery or emergency situations, when the existing communications infrastructure is destroyed or disabled [3]. Mobile Ad hoc Networks are self-organized, temporal networks which consist of a set of wireless nodes. The nodes can move in an arbitrary manner and work as its own opinions. Nodes communicate with each other by forming a multi-hop radio network and maintaining connectivity in a decentralized manner. Each node in MANETs plays both the roles of routers and terminals. Such devices can communicate with another device that is immediately within their radio range or one that is outside their radio range not relying on access point [4]. A mobile ad hoc network is self-organizing, self-discipline and self-adaptive. The main characteristics of mobile ad hoc network are:  Lack of Infrastructure: (Dynamic topology) since nodes in the network can move arbitrarily, the topology of the network also changes.  Limitations on the Bandwidth: The bandwidth of the link is constrained and the capacity of the network is also variable tremendously [5]. Because of the dynamic topology, the output of each relay node will vary with the time and then the link capacity will change with the link change.  Power considerations: it is a serious factor. Because of the mobility characteristic of the network, devices use battery as their power supply. As a result, the advanced power conservation techniques are very necessary in designing a system [2].  Security Precautions: The security is limited in physical aspect. The mobile network is easier to be attacked than the fixed network. Overcoming the weakness in security and the new security trouble in wireless network is on demand [6].

Page 1

A. A. Salama et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 01-07

A side effect of the flexibility is the ease with which a node can join or leave a MANET. Lack of any fixed physical and, sometimes, administrative infrastructure in these networks makes the task of securing these networks extremely challenging [7]. In MANETs it is very important to address the security issues related to the dynamically changing topology of the MANET [8], these issues may be defined as: 1- Confidentiality. The primary confidentiality threat in the context of MANET is to the privacy of the information being transmitted between nodes, which lead to a secondary privacy threat to information such as the network topology, geographical location, etc. 2- Integrity. The integrity of data over a network depends on all nodes in the network. Therefore threats to integrity are those which either introduce incorrect information or alter existing information. 3- Availability. This is defined as access information at all times upon demand. If a mobile node exists, then any node should be able to get information when they require it. Related to this, a node should be able to carry out normal operations without excessive interference caused by the routing protocol or security. 4- Authorization. An unauthorized node is one which is not allowed to have access to information, or is not authorized to participate in the ad hoc network. There is no assumption that there is an explicit and formal protocol, simply an abstract notion of authorization. However, formal identity authentication is a very important security requirement, needed to provide access control services within the ad hoc network. 5- Dependability and reliability. One of the most common applications for ad hoc networks is in emergency situations when the use of wired infrastructure is infeasible. Hence, MANET must be reliable, and emergency procedures may be required. For example, if a routing table becomes full due to memory constraints, a reactive protocol should still be able to find an emergency solution. 6- Accountability. This will be required so that any actions affecting security can be selectively logged and protected, allowing for appropriate reaction against attacks. The misbehaviors demonstrated by different types of nodes will need to be detected, if not prevented. Event logging will also help provide nonrepudiation, preventing a node from repudiating involvement in a security violation [9]. 7- Non-repudiation Ensures that the origin of a message cannot deny having sent the message. Intuitionistic fuzzy sets can be viewed as a generalization of fuzzy sets that may better model imperfect information which is omnipresent in any conscious decision making. The rest of this paper is organized as follows; some backgrounds are given in section 1. Section 2 provides the proposed security mechanism. A comparison of the proposed mechanism with some of the current security mechanisms is provided in section 3. Section 4 provides the conclusions and future work. A. Intuitionistic Fuzzy Sets A fuzzy set is a nebular collection of elements from a universe M described by and identified with a (membership) function A: M￫ [0, 1] [10]. An intuitionistic fuzzy set is instead a nebular collection of elements from M identified with a pair

( A, Ad ), where

A, Ad: M￫[0, 1] and ∀x∊M: A( x) + Ad ( x) ≤ 1. one interprets A as a membership function: A( x) is a degree of membership of x in the intuitionistic fuzzy set , whereas Ad, a function dual to A, is understood as a non-membership function, i.e. Ad( x) does express a degree of non-membership of x in that intuitionistic fuzzy set. Finally the term called the degree of hesitation whether or not x is in (x) = 1- ( A+ Ad ) [15].

in the following equation is

B. Public Key Security The distinctive technique used in public key cryptography is the use of asymmetric key algorithms, where the key used to encrypt a message, not the same as the key used to decrypt it. Each user has a pair of cryptographic keys - a public encryption key and a private decryption key [11]. The provision of public key cryptography is widely distributed, while the private-decryption key is known only to the recipient. Messages are encrypted with the recipient's public key and can only be decrypted with the corresponding private key. The keys are mathematically related, but the parameters are chosen so that the determination of the private key of the public key is prohibitively expensive. The discovery of algorithms that can produce pairs of public / private key revolutionized the practice of cryptography in principle in mid-1970.In contrast, symmetric key algorithms, variations of which have been used for thousands of years, uses a single secret key - that should be shared and kept private by the sender and receiver - for encryption and decryption. To use a symmetric encryption scheme, the sender and receiver must share the key securely in advance. Because symmetric key algorithms are almost always much less computationally intensive, it is common to exchange a

Page 2

A. A. Salama et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 01-07

key using a key exchange algorithm and transmit data using that key and symmetric key algorithm [12]. Family PGP and SSL / TLS schemes do this, for example, and therefore speak of hybrid cryptosystem.  The two main branches of public key cryptography are: Public Key Encryption: a message encrypted with the recipient's public key can be decrypted by anyone except a holder of the corresponding private key - presumably this will be the owner of that key and the person associated with the public key used. This is used for confidentiality [13].  Digital signatures (Authentication): a signed message with the sender's private key can be verified by anyone with access to the sender's public key, which shows that the sender had access to the private key (and therefore likely to be the person associated with the public key used), and part of the message has not been tampered with. On the question of authenticity, see also the summary of the message [14]. The main idea behind public-key (or asymmetric) cryptosystems is the following: One entity has (in contrast to symmetric cryptosystems) a pair of keys which are called the private key and the public key. These two parts of the key pair are always related in some mathematical sense. As for using them, the owner of such a key pair may publish her public key, but it is crucial that she keeps the private key only for herself. Let (sk, pk) be such a key pair where sk is the Secret private Key for node (A) and pk is the corresponding public key [15]. If a second node wants to securely send a message to (A) it computes: C = encrypt(M, pk) where encrypt denotes the so-called encryption function which is also publicly known as shown in Figure 1.

Figure 1: Asymmetric Key encryption / decryption This function is a one-way function with a trap-door. In other words, the trap-door allows for the creation of the secret key sk which in turn enables Alice to easily invert the encryption function. We call C the ciphertext. Obtaining M from C can be done easily using the (publicly known) decryption function decrypt and A’s private key (sk). On the other hand, it is much harder to decrypt without having any knowledge of the private key. As already mentioned, the great advantage of this approach is that no secure key exchange is necessary before a message is transmitted [16]. II.

The proposed model for security

In this section, a Security algorithm applied to MANETs is presented. This algorithm may be viewed as a two stages: first an intuitionistic fuzzy model to decide the key length for the current session. Then the key distribution between nodes in MANET both stages are illustrated in the rest of this section. A. Intuitionistic fuzzy model (Key Size Determination Function) The security offered by the algorithm is based on the difficulty of discovering the secret key through a brute force attack. Mobile Status (MS) Security Level is the correlative factor being analyzed with three considerations: 1- The longer the password, harder to withstand a severe attack of brute force. In this research the key lengths from 16 to 512 are assumed 2- The quickest way to change passwords, more secure the mobile host. It is more difficult to decipher the key to a shorter time. A mobile host to change the secret key is often safer than a mobile host using a constant secret key. 3- The neighbor hosts the mobile host has, the more potential attacker. I.e. the possibility of attack is greater. There are many other factors affecting the safety of mobile hosts, such as bandwidth. The security level of mobile hosts is a function with multiple variables and affected more than one condition. Here a intuitionistic fuzzy logic system is defined. Inputs of the intuitionistic fuzzy logic system are the frequency of changing keys (f ) and the number of neighbor hosts (n). Output of the intuitionistic fuzzy logic system is the Security-Level of MS. It is assumed that the three factors are independent with each other. The relationship of them is as follows: Formula 1 It means that the Security-Level of MH is in direct proportion to the length of the key and the frequency of changing keys, in inverse proportion to the number of neighbor hosts. The S value is updated by the intuitionistic fuzzy logic system. When the key length is short, the Security-Level of MH should be low; otherwise the Security-Level of MS should be high.

Page 3

A. A. Salama et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 01-07



The first input parameter to the intuitionistic fuzzy variable “the number of neighbor hosts” has three intuitionistic fuzzy sets—few, normal and many. The membership function of n is illustrated in Figure 2. Number of neighbors



Figure 2: Membership function of intuitionistic fuzzy variable n. The input intuitionistic fuzzy variable “the frequency of changing keys” has two intuitionistic fuzzy sets—slow and fast. The membership functions of f is showed in formulation (2)

Formula 2 The output intuitionistic fuzzy variable “the Security-Level of MS” has five intuitionistic fuzzy sets containing the set and its complementary set. These sets are(lowest, low, normal, high and highest). It should be noted that modifying the membership functions will change the sensitivity of the intuitionistic fuzzy logic system’s output to its inputs. Also increasing the number of intuitionistic fuzzy sets of the variables will provide better sensitivity control but also increases computational complexity of the system. Table 1 show the rules used in the intuitionistic fuzzy logic system. Table 1: The Intuitionistic fuzzy system rules Input Output F N S Slow Few (Low , ~Low) Slow Normal (Lowest , ~Lowest) Slow Many (Lowest , ~ Lowest) Fast Few (Normal , ~ Normal) Fast Normal (Low , ~ Low) Fast Many (Low , ~Low) Slow Few (High , ~High) Slow Normal (Normal , ~ Normal) Slow Many (Low , ~ Low) Fast Few (Highest, ~ Highest) Fast Normal (High , ~ High) Fast Many (High , ~ High) The output of that system determines the number of bits used and the security level required for the current situation varying the number of bits between 16 and 256 bits. This determination is based on the IFS analysis whish passes the two parameters ( A, Ad ) then based on that analysis the system decides the accurate key size in each situation 

B. key distribution Once the intuitionistic fuzzy function has decided the length of the session key based on its criteria the problem of key creation and distribution arises. The nature of NANET poses great challenges due to the lake of infrastructure and control over the network. To overcome such problems the use of PK scheme is used to distribute the key under the assumption that one node (let us say the first node that originates the network) is responsible for the creation of session keys. If that node is going to leave the network it must transfer the process of key creation to another trusted node in the network. 1- Each node sends a message (Session Key Request SKR) encrypted with its private key (that message contains a key request and a timer) to the key creator node which owns a table that contains the public key for each node in the network. Figure 3 (a) where the direction of the arrow’s head denotes the private key used encryption is the originating node. 2- The key creator node simply decrypts the message and retrieves the request and the timer with one of the following scenarios occurs: a. The timer was expired or the message is unreadable the message is neglected. b. The timer is valid and the decryption of the message using the corresponding Public Key gives a readable request. The key creator node sends a message to that node containing the current session key. That message is encrypted two times first using the key creator’s Private key(for authentication) then using the destination’s public key Figure 3 (b). Where the direction of the arrow’s head denotes the private key used encryption is the trusted node then with the destination node’s Public Key.

Page 4

A. A. Salama et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 01-07

3- Any time the intuitionistic fuzzy model reports that the network condition changes; the key creator node sends a jamming message for every node currently in the network asking them to send a key request message. 4- Any authenticated node (including the Trusted node) on the network knowing the current session key can send messages either to every node or to a single node on the network, simply by encrypting the message using the current session key. . Figure 3 key distribution: (a) SK Request (b )SK Response Node A

Node A

Node B

Trusted Node

Node C

Node D

Node C

Node D

(a) III.

(b)

Experimental Results

In this research a new security algorithm for MANETs is presented, this algorithm is based on the idea of periodically changing the encryption key thus make it harder for any attacker to track that changing key. The algorithm is divided into two stages key size determination function and key distribution. In this section the set of experimental results for the attempts to decide the way for creating a more secured MANETs. These experiments are clarified. A. Intuitionistic fuzzy vs. Non-Intuitionistic fuzzy Key size determination function: The first type of experiments had taken place to decide the key size for the encryption process. To accomplish this job the ordinary mechanism of KNN is used as a non-intuitionistic fuzzy technique. Given the same parameters passed to the intuitionistic fuzzy and the non-intuitionistic fuzzy function the performance is measured with evaluation criteria are the average security-level and the key creation time. The performance criteria are demonstrated in the following sections: A.1 The Average security-level: Average security level is measured for both techniques as the corresponding key provided how much strength given the number of nodes, the results are scaled from 0 to 5 these results are shown in table 2 and figure 4. Table 2 ASL of intuitionistic fuzzy vs. non-intuitionistic fuzzy classification No. nodes

100

125

150

175

200

225

250

Non-Intuitionistic fuzzy Classification Intuitionistic fuzzy Classification

2.6

2.1

2.5

2.2

1.5

1.7

1.4

2.3

1.5

3.4

3.6

3.8

3.9

Figure 4: average security-level vs the number of mobile nodes 5 4 3 Non-Fuzzy

1 25 50 75 100 125 150 175 200 225 250

0 Figure 4 and table 2 shows the average security level with the number of mobile nodes between 25 and 250. As shown in the figure and the table, the average security-level of the Intuitionistic fuzzy Classifier (FC) is much higher than the average security-level of the non-intuitionistic fuzzy classifier, especially for many mobile nodes. This is an expected result since the intuitionistic fuzzy classifier adapts its self upon the whole set of criteria. A.2 The key creation time: The time required to generate the key in both cases are measured, the results are scaled from 0 to 1 and are shown in table 3 and figure5

Page 5

A. A. Salama et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 01-07

Table 3: KCR of intuitionistic fuzzy vs. non-intuitionistic fuzzy classifiers No. nodes

100

125

150

175

200

225

250

Non-Intuitionistic fuzzy Classification

0.95

0.93

0.95

0.96

0.93

0.9

0.85

0.92

0.93

0.94

Intuitionistic fuzzy Classification

Figure 5: Key creation time vs the number of mobile nodes. 1 0.95 0.9 Non-Fuzzy 0.85

0.8 25 50 75 100 125 150 175 200 225 250

0.75 Figure 5 and table 3 shows the Key creation time with the number of mobile nodes between 25 and 250. The speed of Key creation is very high (mostly above 0.94) for all two techniques. However, the Non-intuitionistic fuzzy technique has some faster Key creation time than the Intuitionistic fuzzy Classifier, especially with few mobile nodes. The reason is that the smaller the number of nodes with the same amount of calculation the bigger the time taken. B. PKI vs. non-PKI distribution After the Key size had been determined via the Key size determination function the final problem is to distribute that key among nodes on the network. There were two approaches for the key distribution problem either PKI or non-PKI. In this subsection the results of applying PKI and non-PKI techniques is illustrated as applied in terms of security and processing time B.1 Security The PKI presents more overall security than ordinary non-PKI (single key) that is illustrated by applying both techniques over the network and recording the results regarding to the time required for an external attacker to break the session key. Table 4 and figure 6 shows that results under the assumption of using small public-private key pairs. Table 4: security of PKI vs, non-PKI No. nodes Non-PKI

100

125

150

175

200

225

250

0.15

0.2

0.23

0.26

0.3

0.32

0.36

0.4

0.44

0.45

0.85

0.92

0.93

0.94

PKI 0.8

Figure 6: security of PKI vs, non-PKI 1 0.8 0.6

Non-PKI

0.4

PKI

0.2 25 50 75 100 125 150 175 200 225 250

In graph and figure shows the huge difference in the security level provided by the PKI technique over the NonPKI mechanism given the same experimental conditions. B.2 Processing time Another factor had been taken into consideration while developing the model that is time required to process the key and distribute it. Table 5 and figure 7 shows that results under the assumption of using small public-private key pairs.

Page 6

A. A. Salama et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 01-07

Table 5 Processing time of PKI vs. non-PKI No. nodes Non-PKI PKI

100

125

150

175

200

225

250

0.3

0.32

0.35

0.37

0.4

0.44

0.47

0.51

0.55

0.58

0.2

0.35

0.5

0.6

0.68

0.75

0.83

0.87

0.93

0.97

Figure 7: Processing time of PKI vs. non-PKI 1.2 1 0.8 0.6

Non-PKI

0.4

PKI

0.2 25 50 75 100 125 150 175 200 225 250

0 Table 5 and the Figure 7 shows that Non-PKI techniques provides relatively small amount of processing time than PKI this due to the amount of modular arithmetic performed in the PKI mechanisms. However the difference in the processing time is ignored comparing to the security level provided by the PKI under the same conditions IV. Conclusions MANETs require a reliable, efficient, and scalable and most importantly, a secure protocol as they are highly insecure, self-organizing, rapidly deployed and they use dynamic routing. In this paper, we discussed the vulnerable nature of the mobile ad hoc network. Also the security attributes and the various challenges to the security of MANET had been covered. The new security mechanism which combines the advantages of both intuitionistic fuzzy classification and the public key infrastructure had been demonstrated. The advantages of the proposed mechanism comparing to other existing mechanisms had been shown by first comparing the intuitionistic fuzzy to the non-intuitionistic fuzzy classification showing that intuitionistic fuzzy is more adaptable and provides a better response in MANET. Also the PKI is compared to the non-PKI showing that it provides a far better security with a ignored amount of delay. References [1.]

[2.]

[3.] [4.] [5.] [6.] [7.]

[8.] [9.] [10.] [11.] [12.] [13.] [14.] [15.] [16.]

I.m.hanafy, A.a.salama , M.abdelfattah & Y.m.wazery, “AIS MODEL FOR BOTNET DETECTION IN MANET USING FUZZY FUNCTION”, International Journal of Computer Networking, Wireless and Mobile Communications (IJCNWMC) ISSN 2250-1568 ,Vol. 3, Issue 1, Mar 2013, 95-102 Balakrishnan, V. Varadharajan, U. K. Tupakula, and P.Lucs, "Trust Integrated Cooperation Architecture for Mobile Ad-hoc Networks". Proceedings of 4th IEEE International Symposium on Wireless Communication Systems (ISWCS 2007), Trondheim, Norway, 2007. AW. Stallings; “Cryptography and Network Security – Principles and Practice”, 9th Edition; Prentice Hall 2010 Dr.A.Rajaram, S.Vaithiya lingam. “Distributed Adaptive Clustering Algorithm for Improving Data Accessibility in MANET”. IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 4, No 1, July 2011 S. Balachandran, D. Dasgupta and L. Wang. “ Hybrid Approach for Misbehavior Detection in Wireless Ad-Hoc Networks”. Published in Symposium on Information Assurance, New York, June 14-15, 2006. A.Rajaram, S.Palaniswami .” THE MODIFIED SECURITY SCHEME FOR DATA INTEGRITY IN MANET”. International Journal of Engineering Science and Technology. Vol. 2(7), 2010, 3111-3119 C Balakrishnan, V. Varadharajan, U. K. Tupakula, and P.Lucs, "Trust Integrated Cooperation Architecture for Mobile Ad-hoc Networks". Proceedings of 4th IEEE International Symposium on Wireless Communication Systems (ISWCS 2007), Trondheim, Norway, 2007. A. Srinivasan, J. Teitelbaum, H. Liang, J. Wu, and M. Cardei, "Reputation and Trust-Based Systems for Ad-hoc and Sensor Networks," Algorithms and Protocols for Wireless Ad-hoc and Sensor Networks, A. Boukerche (ed.), Wiley & Sons, 2011. K.Seshadri Ramana et al.” Trust Based Security Routing in Mobile Adhoc Networks”, (IJCSE) International Journal on Computer Science and Engineering Vol. 02, No. 02, 2010, pp 259-263. Yan L. Sun, Wei Yu, ”Information Theoretic Framework of Trust Modeling and Evaluation for Ad Hoc Networks”, 2006 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 24, NO. 2, FEBRUARY, pp305-317 Er. Banita Chadhaa, Er. Zatin Gupta,” Security Architecture for Mobile Adhoc Networks” (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 9, 2011, pp 101 – 104 F. L. Bauer. Decrypted Secrets: Methods and Maxims of Cryptology. Springer, Secaucus, NJ, USA, 9th edition, 2009. ISBN:3540668713 Dabrowski J. and Kubale M., Computer Experiments with a Parallel Clonal Selection Algorithm for the Graph Coloring Problem. IEEE International Symposium on Parallel and Distributed Processing (IPDPS 2008), 14-18 April, Miami, FL, USA, pp.1-6. Rajaram A and Palaniswami S, “A Trust-Based Cross-Layer Security Protocol for Mobile Ad hoc Networks”, International Journal of Computer Science and Information Security, Vol. 6, No. 1,p.p 165 – 172, 2009. Reza Azarderskhsh, Arash Reyhani-Masoleh.” Secure Clustering and Symmetric Key Establishment in HeterogeneousWireless Sensor Networks”. EURASIP Journal onWireless Communications and Networking .2011, K. Ren, S. Yu, W. Lou, and Y. Zhang, “Multi-user broadcast authentication in wireless sensor networks,” IEEE Transactions on Vehicular Technology, vol. 58, no. 8, pp. 4554–4564, 2009.

Page 7

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Automatic Detection and Classification of Glioma Tumors using Statistical Features Ananda Resmi S.1, Tessamma Thomas2 Department of Electronics and Communication Engineering, College of Engineering Perumon Kollam, Kerala, INDIA 2 Department of Electronics, Cochin University of Science and Technology Kochi-22, Kerala, INDIA ______________________________________________________________________________________ Abstract: The characterization and grading of glioma tumors, via image derived features, for diagnosis, prognosis, and treatment response has been an active research area in medical image computing. This paper presents a novel method for automatic detection and classification of glioma from conventional T2 weighted MR images. Automatic detection of the tumor was established using newly developed method called Adaptive Gray level Algebraic set Segmentation Algorithm (AGASA).Statistical Features were extracted from the detected tumor texture using first order statistics and gray level co-occurrence matrix (GLCM) based second order statistical methods. Statistical significance of the features was determined by t-test and its corresponding pvalue. A decision system was developed for the grade detection of glioma using these selected features and its pvalue. The detection performance of the decision system was validated using the receiver operating characteristic (ROC) curve. The diagnosis and grading of glioma using this non-invasive method can contribute promising results in medical image computing. 1

Keywords: Glioma; Automatic Detection; Texture; GLCM; t-test; p-value; Feature extraction; Classification ________________________________________________________________________________________ I. Introduction Gliomas are the most frequent primary brain tumors that originate in glial cells. Glial cells are the building-block cells of the connective, or supportive tissue in the central nervous system (CNS) [1], [2]. According to World Health Organization (WHO), gliomas are classified into four grades that reflect the degree of malignancy. Grades I and II are considered low-grade and grades III and IV are considered as high-grade. Grades I and II are the slowest-growing and least malignant, Grade III tumors are considered malignant and grow at a moderate rate. Grade IV tumors, such as Glioblastoma multiforme, are the fastest growing and the most malignant primary brain tumors [1], [3]. Classification of glioma tumors is important for clinical understanding of tumor biology, clinical response and for assessing overall prognosis with brain tumors. Conventional MR imaging is the standard technique for diagnosis, treatment planning, and monitoring of CNS lesions, with superior sensitivity compared to alternative modalities [4]. It is routinely used for the noninvasive assessment of brain tumors, but its ability to define the tumor type and grade of gliomas is limited [5]. A biopsy and surgical resection is usually required to establish the diagnosis and subtype of a brain tumor after conventional MR imaging, but variations in tissue sampling may produce erroneous result during biopsy [6]. In this work, only conventional T2-weighted MR images are considered and this modality highlight tissues with higher concentration of water in which border definition and tumor heterogeneity are best observed [7]. Most of the segmentation technique in literature such as fuzzy c-means clustering [8], [9], Markov random fields [10], [11] level set method [12] model based techniques [13] are time consuming or complex or need human intervention. Accurate segmentation of glioma is also very important in this work because entire portion of tumor texture is considered for classification of tumor. Hence a new method for accurate detection glioma tumor is described here. Texture features have proved useful in differentiating normal and abnormal tissues [14] in different organs using different types of imaging modalities. Texture analysis is very important in the brain tumor detection, as it is difficult to differentiate between various types of tumor tissues using shape feature alone [15]. 2D textural features have been previously employed for MRI brain tumor characterization and pattern recognition systems [16]. Classifications of primary and secondary brain tumors using first order and second order statistics from MRIs [17] have also been developed. Statistical analysis of textures from brain CT images for quantifying

Page 8

Ananda Resmi S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 08-14

tumor heterogeneity and thereby differentiate high and low grade glioma are there in literature [18]. Texture analysis using statistical quantification has proposed in literature for differentiating glioneuronal tumors as a subclass of grade III and IV malignant gliomas [19] Gray level co-occurrence (GLCM) based texture analysis is widely used in the detection of breast cancer in mammograms [20] detection of abnormal liver in CT images [21] and detection of primary and secondary tumors [17] in brain MRIs. Several approaches are developed in the literature for classification and grade detection of glioma tumors. Classification of glioma from metastatic and grading of glioma from conventional MRI and perfusion MRI, using support vector machines (SVM) [6], Artificial Neural Network (ANN) [22] and linear discriminant analysis (LDA) [23] is cited in literature. The features used for their study were tumor shape, intensity characteristics as well as rotation invariant Gabor texture features. Determination of degree of malignancy of glioma using SVM is [24], [25] also developed in the literature. The degree of malignancy was determined in their work using the features from clinical data before operation and findings from conventional T1 and T2 weighted MRI such as age, shape, gender etc. The objective of this work is automatic detection and classification of low and high grade glioma from T2 weighted brain MRI. The framework of the method consists of detection of region of interest (ROI) using Adaptive Gray level Algebraic set Segmentation Algorithm (AGASA), feature extraction from detected tumor texture based on first order and GLCM based second order statistics, feature selection using t-test and its corresponding p-value, and classification of low and high grade glioma, based on p-values of selected features, training and performance evaluation of results using receiver operating characteristic curve (ROC). II. Materials and Methods The block diagram for the entire procedure is shown in Fig.1. The steps involved were image preprocessing, detection of glioma, feature extraction process, feature selection and classification.

MRI Data Set

Texture Feature extraction & selection

Detection of Tumor

Image preprocessing

Classificati on

Segmente d ROI

Preprocessed image

Fig.1 Block diagram of the method

A. Image Database The study population comprised of T2 weighted axial MRI data sets (of 135 glioma patients (87 men, 48 women; Age 18-78 years) for detection and classification. Out of this, 75 were of high grade glioma and 60 sets were of low grade.

Fig.2 Shows T2 weighted images of low grade and high e grade glioma tumors All selected image dataset consisting of single glioma was considered for detection and classification. All patients underwent biopsy or surgical resection of the tumor with histopathological diagnosis. MR images were collected from the department of Radiology in the Sree Chitra Institute of Medical Sciences and Technology (SCIMST) and Regional Cancer Centre Trivandrum, India. The images were gray scale images and which were acquired before contrast enhancement. Fig.2 shows the examples for T2 weighted images of low and high grade tumors.

Page 9

Ananda Resmi S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 08-14

B. Detection of Tumor A novel method is discussed here for accurate detection of tumor tissues using Adaptive Gray level Algebraic set filtering Algorithm (AGASA). Fig.3 shows the detection of glioma. The method mainly involves repeated mathematical morphology based operations using different structuring elements [26], thresholding and masking techniques. As the first part of segmentation procedure MR Image is complimented and dilated using square shaped SE. The complemented image is subtracted from the dilated image. The subtraction of dilated image from the complemented image is done for reducing noise artefacts and partial volume effect present in the image and enhancing tumor. MR Image

Image complementation

Morphological dilation

Image subtraction

Image filtering

Morphological opening and closing

Image thresholding and masking

Detected ROI

Fig.3 The flow chart for Detection of glioma The resulting image undergoes spatial domain filtering by correlation method for enhancing regions which are similar to the filter templates. The filtered output is again dilated with a square shaped SE. The main challenge in the tumor segmentation procedure is that usually tumor boundaries will not be clearly defined from the other regions and tumors may have heterogeneous borders and will have infiltrating nature. This boundary intrusions and protrusions are clearly visible after dilation. The main disadvantage of morphological dilation is over segmentation [27], this can be reduced to a greater extent using morphological opening operation with a disc shaped SE of suitable radius. After the opening operation, the output image undergoes closing operation. The combination of opening followed by closing or closing followed by opening can suppress noise sufficiently [27]. The tumor boundary and region of the resulting image is visually enhanced. The resultant image is thresholded by a specific threshold level to obtain the segmented ROI. The binary image thus obtained was masked with the original normalized image in order to obtain the original gray level image of the corresponding ROI. The detected tumor is validated with manual ground truth [26] C. Feature description The detected tumor was considered for texture analysis. A set of textural descriptors was calculated for each ROI in the training set, using first order statistics and GLCM based second order statistics []. 1) First-order statistics For any region of interest (ROI), the mean (average Intensity) and the standard deviation (average contrast) of the gray level values in the region can be used to measure the spread of gray level values of the pixels within that region (Histogram). One class of such measures is based on statistical moments. Here statistical moments such as mean, standard deviation, entropy, kurtosis and skewness are calculated from the segmented ROI. Entropy indicates a measure of irregularity, kurtosis indicates a measure of peakedness and skewness indicates a measure of asymmetry [28]. Five descriptors were computed using first order statistics. Quantification using these first order statistical features, one can determine the given texture is coarse or fine.

Page 10

Ananda Resmi S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 08-14

Second order statistics GLCM is a widely used tool for analyzing statistical textural properties of different types of tissues in biomedical imaging. From the co-occurrence model, 10 Halarick descriptors are calculated in order to quantify the spatial dependence of gray level values. These descriptors are computed from the co-occurrence matrices of size [16x16], which is constructed at with an inter pixel distance of d = 1 and for a pixel direction Î¸ = 0Â° [28]. The main texture descriptors derived from GLCM are, Correlation, Contrast, Energy, Entropy, Cluster Prominence, and Cluster Shade of gray level values and the other descriptors are relative values of these features. Contrast evaluates the amount of local intensity variations present in an image, and energy is the sum of squared elements in GLCM. It is also a measure of uniformity and angular second moment. The Correlation of a texture depicts the linear dependency of gray levels on neighboring pixels [29]. D. Feature selection Feature selection and feature set formulation are very important, because selected features must be sufficiently discriminating and suitably adapted for the application, since they fundamentally impact the resulting quality of the detection system. Fifteen feature descriptors were extracted from the first order statistics and GLCM methods. The t-test (test statistic) checks whether the means of two groups are statistically different or in other words it determines, the two dataset come from the same population or different population [29]. The p-value is associated with a t-test. It is the level of marginal significance within a statistical hypothesis test, representing the probability of the occurrence of a given event. The smaller the p-value, the more strongly the test rejects the null hypothesis, that is, the hypothesis being tested. The t- test for extracted 5 first order statistical features and 7 GLCM features of low and high grade glioma are performed. The statistical significance of each feature for the two datasets is estimated using its p-value. Using the statistically significant features a decision system is developed for classification of low and high grade glioma. This decision system does not make any assumptions about the distribution of data. The training set was used to build the decision system, while test set was used to estimate the accuracy of the system. The training and testing are based on the p-value of each feature in the feature set. E. Performance evaluation The accuracy and performance of detection system can be analyzed using four parameters [30], which are false positive (FP), false negative (FN), true positive (TP), and true negative (TN). For evaluating the accuracy of detection, specificity and sensitivity of detection have to be considered. Sensitivity (eqn.(1)) and specificity (eqn. (2)) are two important parameters which indicate the presence or absence of the disease.

In particular, Sensitivity is the percentage of correctly diagnosing high grade glioma and the result is also positive. Specificity indicates false positive rate (FPR). Thus, it is a measure of the probability of correctly distinguishing the absence of high grade glioma and the result is negative. The ROC analysis is done for assessing the performance. It is a graphical plot of the sensitivity against specificity for a binary classifier system at different operating points. For a perfect classifier, ROC curve will pass through upper left corner (0, 1) of the ROC space.. III. Results and Discussion A. Detection of Tumor This section presents the results obtained from the automatic segmentation using spatial domain filtering techniques on T2 weighted axial MR images of 135 MR image datasets. Usually there are only four to eight slices which contain maximum gray levelbinary tumor information in a data set. segmented tumor

Figure 4 Segmentation procedures from low grade glioma tumor from a T2 weighted MR image (a) original image (b) Image after thresholding (c) Segmented gray level tumor

Page 11

Ananda Resmi S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 08-14

Fig. 4a to Fig. 4c shows the detection of images. The final output of a segmentation process is a binary image as shown in Fig. 4b. In order to retrieve the texture information, the segmented image is masked with the original image. Finally, a binary tumor mask was obtained after removing all background details using detection algorithm. This tumor mask is multiplied with the gray level image for obtaining gray level tumor. The extracted gray level tumor will be used for texture analysis. B. Feature extraction and feature selection Table.1 portrays the ranges of values for first order statistical descriptors, such as intensity, standard deviation, entropy, kurtosis, and skewness for the two grades of glioma tumors and its respective p-values. These ranges of values determined from the statistical quantification of 135 MRI datasets of segmented ROIs (75-low grade, 60- low grade). The t-test was performed for the first order statistical features and its statistical significance was tested for a confidence interval of 0.05 and the corresponding p-value was observed. From the Table 1, it can be observed that p-value for all the five features were <<0.001. If the p-value of two data sets are <<0.001 indicates that mean of two data sets are different or these datasets are statistically different and it is very much less than the confidence interval and the test strongly rejects the null hypothesis. This proved the effectiveness selecting these features for detection of low and high grade glioma. Table1. The ranges of values for first order statistical features for low and high grade glioma (training set) and its p-value First order statistical features

High Grade

Low grade

P-value;

Intensity

190-240

70-160

p<<0.001

Std.dev.

90-150

10-60

p<<0.001

Kurtosis

115-152

1.5-12

p<<0.001

Entropy

6.5-15

0.5-5.2

p<<0.001

Skewness

8-25

0.2-3.0

p<<0.001

These statistical descriptors yield characterization of high grade glioma texture as coarse texture. Usually coarse textures show heterogeneous behaviour. Entropy is a measure of randomness. Highly malignant glioma (grade IV) tumors contain heterogeneous tumors [31]. As malignancy increases heterogeneity is also increasing. Intensity, Standard Deviation, Third Moment (Skewness), Kurtosis, and Entropy are low for low grade glioma. Low grade gliomas have smooth textures when compared with high grade. This proved the effectiveness of first order statistical descriptors theoretically and these well differentiated features were selected for decision system for detection. Table 2. The ranges of values for second order (GLCM) statistical features for low and high grade glioma (135 patients) and its p-value GLCM features

High Grade

Low grade

P-value

Auto Correlation

40-75

10-30

p<<0.001

Contrast

25-55

5.0-16

p<<0.001

Cluster prominence

1300-1700

425-650

p<<0.001

Cluster Shade

100-200

20-80

p<<0.001

Entropy

6-15

0.2-1.5

p<<0.001

Dissimilarity

50-250

0.5-10

p<<0.001

Energy

0.2-1.5

3-15

p<<0.001

Table 2 illustrates the ranges of values for GLCM based second order statistical features and its corresponding p-values. From the Table 2, it can be noticed that, GLCM based texture descriptors for high grade glioma: dissimilarity, entropy, contrast, cluster shade, cluster prominence and auto correlation are high whereas energy is low for high grade glioma texture. These features are well enough for characterizing textural properties of a subject. Contrast, Dissimilarity, Entropy and energy are measures of non-uniformity or randomness of a texture and it is strongly correlated with texture heterogeneity [25]. From the p- values it was

Page 12

Ananda Resmi S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 08-14

observed that mean of data sets are well differentiated and the two data sets are statistically different. Hence these features were selected for classification of low and high grade glioma . C. Classification of low grade and high Glioma tumors and evaluation of results Detection was done using on twelve features using selected textural descriptors. Based on p-values of the features, a decision system was developed. Statistical significance of each feature was tested according to the test rejects or accepts the null hypothesis and grade of glioma is determined. If the test is rejected decision system will detect the ROI for the decision level is true are considered as high grade or otherwise it is a low grade tumor. Out of the 135 patient image dataset, 85 data sets were used for training purposes and 50 data sets (30-high grade glioma, 20-low grade glioma) were used for test purposes. The performance of detection system was evaluated using ROC curve (Fig.5). The graph depicts the trade-off between the true-positive and falsepositive rates. The sensitivity and specificity of the detection system is 99.03% and 99.53% respectively.

Conclusion Fig.5 shows the ROC curve for detection system Sensitivity -99.03%, specificity-99.53%, Performance of detection-Excellent test The method in this work showed better performance than the other existing methods [6], [18], [35]. In this work, the features selected are well discriminated between two grades and hence a decision system is sufficient for detection process. As per citations [31] tumor heterogeneity and degree of malignancy is directly related and well established using texture analysis. IV. Conclusion A novel method for automatic detection and classification of low and high grade glioma from conventional MR images were presented in this paper. The axial slices of T2 weighted MRI were considered for the method. Adaptive Gray level Algebraic set Algorithm was developed for accurate detection of glioma. It is important to detect of tumor texture accurately, because entire portion is considered for further analysis and classification. Statistical texture analysis of tumor texture was done using first order and GLCM based second order statistics. Statistical significance of these features for low and high grade glioma was determined using ttest and its corresponding p-value was computed. Based on these p-values of selected features a decision system was developed for grade detection. The performance of the decision system was evaluated. The sensitivity and specificity of the detection system are 99.03% and 99.53% respectively. This method is very simple and accurate than the other existing methods. Along with the statistical features by incorporating histopathological properties, edema properties, tumor shape etc., more sophisticated and robust system could be developed for detecting all grades and sub types of glioma. References [1]

E.Mandonnet, Capelle, L. and U.Duffau, ‘Extension of paralimbic low grade gliomas: toward an anatomical detection based on white matter invasion patterns’, Journal of Nuero-Oncology, Vol.78 No.2 , 2006, pp.179-185.

[2]

D.Schiff, ‘Low grade Astrocytomas’ 2007, An article in American Brain Tumor Association

[3]

A. Michotte, B. Neyns, C.Chaskis, et.al , ‘Neuropathological and molecular aspects of low-grade and high grade gliomas’, Acta nuerol. beig.,Vol. 104 , 2004, pp.148-153.

[4]

M. Essig,. N.Anzalone, S. E. Combs, et.al (2012) ‘Imaging of Neoplastic Central Nervous System Lesions: Review and Recommendations for Current Practice’, American Journal of Neuroradiology Vol.33, pp.803-17.

[5]

W. Bian, S. K Inas , M. L.Janine et.al ‘Multiparametric Characterization of Grade 2 Glioma using magnetic Resonance Spectroscopic, perfusion, and Diffusion Imaging’, Translational Oncology, Vol. 2 No.4. 2009 , pp.271-280

Page 13

Ananda Resmi S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 08-14

[6]

E. I. Zacharaki, S.Wang, S Chawla,.et.al Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme, Magn Reson. Med. Vol.62 , 2009 , pp.1609–1618.

[7]

S. cha, , ‘Update on Brain Tumor Imaging: from Anatomy to Physiology’, American Journal of Neuroradiology 2006,pp.475-87

[8]

D. L Pham,. and J. L Prince,.’ An Adaptive Fuzzy C-Means Algorithm for Image Segmentation in the Presence of Intensity Inhomogeneities’ , Pattern Recognition Letters, vol.20, 1998, pp.57-68

[9]

A. Kouhi, H.Seyedarabi, A.Aghagolzadeh, ‘ A Modified FCM Clustering Algorithm for MRI brain image Segmentation’. Proceedings of Machine Vision and Image Processing 2011, pp.1-5

Vol. 27,

[10] R.S Alomari,. K.Suryaprakash, C Vipin, ‘Segmentation of the Liver from Abdominal CT Using Markov Random Field model and GVF Snakes, International Conference on Complex’, Intelligent and Software Intensive Systems IEEE Computer Society 2008, pp.293-298. [11] M.P. Katia, , P. Paolo, L. Brambilla, et.al , ‘ITAC volume assessment through a Gaussian hidden Markov random field model-based algorithm’. Proceedings of 30th Annual International IEEE EMBS conference 2008. [12] P.Chandra Barman, S.Miah, S.Chandra ,’MRI Image Segmentation Using Level Set Method and Implement a Medical Diagnosis System’, Computer Science & Engineering: An International Journal (CSEIJ), Vol.1, No.5, 2011, pp.1-10 [13] O. Colliot, J.Atif, I.Bloch, ‘3D Brain Tumor Segmentation in MRI Using Fuzzy Classification, Symmetry Analysis and Spatially Constrained Deformable Models’, Fuzzy Sets and Systems,vol.160, 2009, pp.1457–1473 [14] D.S. Raicu, J. D Furst, Channin, et . al’ A Texture Dictionary for Human Organs Tissue Classifications’, Proceedings of the 8th world Multi conference on Systemic, Cybernetics and Informatics, 2004, pp. 18-21. [15] [15] N.Sharma, Ray, A.K. Sharma, et.al’ Segmentation and Detection of Medical Images using Texture-Primitive Features: Application of BAM-Type Artificial Neural Network’, Journal of Medical Physics Vol. 33 No.3, 2008, pp.119-26. [16] V. S Vyas and P. Rege, ‘Automated Texture Analysis with Gabor filter’, GVIP Journal, Vol. 6 No.1, 2006, pp.35-41 [17] P. Georgiadis, D.Cavouras, I.Kalatzis , et.al. ‘ Computer aided Discrimination between Primary and Secondary Brain Tumors on MRI: From 2D to 3D Texture Analysis’, e-Journal of Science & Technology (e-JST) , Vol.8, 2008, pp.9-18. [18] K.Skogen, B Ganeshan , Good, et.al ‘ Imaging heterogeneity in gliomas using texture analysis’, Cancer Imaging. Vol. 11, 2011, pp. 5112-5117 [19] P.A, Eliot, D.Olivie, S.Saikali, et.al ‘ Can Dynamic Contrast-Enhanced Magnetic Resonance Imaging Combined with Texture Analysis Differentiate Malignant Glioneuronal Tumors from Other Glioblastoma?’, Neurology Research International. 2012. [20] A. M. Khuzi, R. Besar, , W. Zaki, et.al ‘ Identification of Masses in Digital Mammogram using Gray Level Co-occurrence Matrices’, Biomedical Imaging Intervention Journal, Vol. 8,No.1, 2009 , e5. [21] Poonguzhali , G. Ravindran,’ Automatic Classification of Focal Lesions in Ultrasound Liver Images using combined Texture Features’, Information Technology Journal. Vol. 7 No.1, 2008, pp. 205-208 [22] A. Kothari, ‘Detection and classification of brain cancer using artificial neural network in MRI Images’, World Journal of Science and Technology Vol.2 No.5, 2012 pp:01-04 [23] D. B. Kadam, S. S. Gade, M. D. Uplane, et.al ‘An Artificial Neural Network Approach for Brain Tumor Detection Based on Characteristics of GLCM Texture Features’, International Journal of Innovations in Engineering and Technology, Vol. 2 No.1, 2013, pp.193-199 [24] V.P Gladis ,Pushpa Rathi, and S.Palani, Linear Discriminant Analysis for Brain Tumor Classification using Feature Selection’, International Journal of Communications and Engineering, Vol.5 No.4, 2012, pp.130-134. [25] Guo-Zheng Li, Jie Yang, Chen-Zhou Ye, et.al , Degree Prediction of Malignancy in Brain Glioma using Support Vector Machines, Computers in Biology and Medicine , Vol.36 No.3, 2006, pp.313-325 [26] A. Resmi, , T. Thomas, B .Thomas ,’ A novel automatic method for extraction of glioma tumour, white matter and grey matter from brain magnetic resonance images’, Biomed Imaging Interv J 2013; 9(2):e21 [27] Gonzalez, R.C. Woods, R.E. ‘A Text Book on Digital Image Processing, 2nd Edition, Pearson Education India. 2002. [28] S. Ananda Resmi, T.Thomas, ‘Texture Description of low grade and high grade Glioma using Statistical features in Brain MRIs’, Int. J. of Recent Trends in Engineering and Technology, Vol. 4, No. 3, Nov 2010, pp.27-33 [29] F. Albregtsen, ‘Statistical Texture Measures Computed from Gray Level Co occurrence Matrices’, Image Processing Laboratory, Department of Informatics, University of Oslo, 2008. [30] L.Hamel, ‘ Model Assessment with ROC curves’. Encyclopedia of Data Warehousing and Mining, IGI Global, 2nd Edition, 2009, pp.1316-1323 [31] S. Dube,. .J.J. Corso, F.Timothy,. et.al.’ Automated MR image processing and analysis of malignant brain tumors: enabling technology for data mining’. AIP Conference Proceedings, 2007, pp. 64-84.

V. Acknowledgments We would like to thank Dr. Bejoy Thomas, M.D., P.D.C.C, Additional Professor, Dept. of Imaging Sciences and Interventional Radiology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Kerala, India for his classifying images and validating this work and for providing MR images Radiology department of Sree Chithra Tirunal Institute medical Sciences, and Regional Cancer Centre, Trivandrum, Kerala, India

Page 14

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Comparative studies of Cerium and Zirconium doped Barium Titanate S.N. Rahman, N. Khatun, S. Islam, N.A. Ahmed Industrial Physics Division, BCSIR Laboratories, Dhaka. Bangladesh Council of Scientific and Industrial Research (BCSIR) Dr. Kudrat-i-Khuda Road, Dhaka-1205, BANGLADESH. __________________________________________________________________________________________ Abstract: Comparative investigations of microstructure and dielectric properties of BaTiO3 ceramics doped with different concentration of CeO2 and ZrO2 have been studied. BaTiO3 samples were prepared using conventional method of solid state sintering at 1250C for four hours. Distinct microstructure was observed with pure and doped BaTiO3. The highest value of dielectric permittivity and the lowest value of loss tangent was found for CeO2 with x=0.1 mole percent. In all investigated samples dielectric constant assumes initially large value at low frequencies (75KHz)and attains constant value at frequency 1MHz. Keywords: BaTiO3, ceramic, dielectric properties, grain size, resistivity. _________________________________________________________________________________________ I. Introduction Ferroelectric BaTiO3 is most commonly used as capacitors, thermistors, varistors and energy converting systems depending on suitable additives. The permittivity of ferroelectric perovskite shows marked change with temperature near transition temperature and frequency (relaxation point). BaTiO3 powder is usually mixed with various types of additives in order to obtain better performance and a good control over grain size and electrical characteristics of ceramics. It has been found that the dielectric properties of polycrystalline BaTiO 3 depend to a great extent, on the grain growth during sintering, on additive type and concentration [1]-[3]. For the application of doped BaTiO3 ceramics as a capacitor, material apart from good density, a high dielectric constant and low loss factor have to be achieved. The main factors that must be taken into account during the processing of ceramics are the homogenization of BaTiO3 powder with a suitable doping concentration, the sintering temperature and the length of time. For capacitor type application, the primary requirements are the high permittivity characteristics and a small temperature variation of permitivity over a wide temperature region. Barium titanate is usually doped with variety of additives such as Bi2O3, CeO2, ZrO2, Nb2O, Dy2O and other oxides, which enhance either sintering rate or densification degree [4],[5] to achieve temperature stable dielectrics. Among the dopants that are used to modify the dielectric and semi conducting properties of BaTiO 3, Zirconium and Cerium are the most useful donor dependent that can be incorporated in Ti+2 sites or Ba+2 sites. Depending on their concentration (Zr or Ce), in BaTiO3 may exhibit semiconducting or insulating properties. Influence of CeO2 on electrical and semiconducting properties of BaTiO3 has been widely investigated [8],[9]. An interesting aspect of CeO2 is the special chemistry of Ce+4 which can readily reduce itself to Ce+3. This property is the basis of several technological applications of CeO +2, such as catalytic supports (for automatic exhaust) and high temperature electrodes. Among the dopant elements, ZrO 2 and CeO2 are used to modify the dielectric and semi conducting properties of BaTiO3, because of their electric and catalytic properties, which are related to oxygen stochiomatry. Studies on the effect of Zr addition to BaTiO3 show that Zr addition shift the curie point and depress the dielectric peak. Zr+4 can substitute Ti+4 ions as isovalent doping. Addition of Zirconia particles into Barium titanate can form a core shell structure[11]. The desired temperature stability in dielectric properties is achieved by the formation of core-shell grain. The aim of this paper is to compare the microstructures and dielectric measurements of Barium Titanate doped with Zr and Ce as a function of ac electric field. II.

Material and Methods

A. Sample preparation The doped specimens used for this investigation were prepared by conventional solid state reaction technique [8]. Analytical grade of BaO, TiO2, ZrO2 and CeO2 were used to make ceramic mixed crystals of Ba(Ti1-xZrx)O3 and Ba1-xCexTiO3 . The disks of 6 mm diameter and approximately of 2 mm thickness were prepared and sintered at 1250°C for four hours in a furnace. Silver paste was used on both sides of the disks for good contact. B. Measurements Furnace (type–ROS 3/20 FNR 7601905 Hochest, temperature 1500°C, Germany) was used for sintering the samples. In this experiment, Agilent 4216 precision LCR meter was used for measuring capacitance(C), resistance(R) and dissipation factor (D) in the frequency range from 75 KHz to 32 MHz.

Page 15

S.N. Rahman et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 15-19

The microstructure and grain size of the as-sintered ceramic samples were determined by a Hitachi S-3400 N variable pressure scanning electron microscope (SEM) equipped with an energy dispersive X-ray spectroscopy (EDS) attachment. Microstructure evaluations were made from as-sintered surface of the samples. The evaluation of grain size and concentration of the elementary particles were monitored by SEM-image and EDS spectra consequently. III.Results and discussions A. Miccrostructural characteristics All the samples sintered at 12500C for four hours were studied by scanning electron microscope. SEM study of the samples revealed that the sintering took place without grain growth. Figure 1: SEM images of (a) Pure BaTiO3, (b) x=0.1 ZrBT, (c) x=0.2 ZrBT, (d) x=0.3 ZrBT, (e) x=0.4 ZrBT (f) x=0.1 CeBT and(g) x=0.2 CeBT (h) x=0.3 CeBT (i) 0.4 CeBT a

Fig.1 shows the microstructure of pure BaTiO3 ceramics which exhibits normal grain growth, a fairly uniform microstructure and homogeneous distribution of BaO and TiO2. The microstructure of BT ceramic doped with ZrO2 and CeO2 at different concentrations exhibit homogeneous grain size. Addition of different concentration of Zr and Ce with pure BT ceramics has influence on dielectric property and microstructure of the as-sintered sample. Figures show the significant change in grain size with small addition of doping element. Average grain sizes of the samples are calculated from the microstructure of the surface taken from Scanning Electron Microscope. Five or six determinations of diameter are made at random and the average is computed from the measurements. Fig. 1(a) exhibits normal grain growth of pure BaTiO3 and shows a fairly uniform microstructure and homogeneous distribution of BaO and TiO2. It is observed that the average grain size is 1.024 ď m in case of pure BaTiO3. Uniformity of microstructure is revealed from EDS analysis. The microstructures of Ce doped BaTiO3 with different concentrations show fairly uniform distribution of the constituent elements with different grain size. Micrograph shows that there are coarse grains as well as small grain. Some Ce and Zr rich regions

Page 16

S.N. Rahman et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 15-19

were detected in microstructure from point shooting analysis of EDS spectra. The additive rich phases are formed in local regions due to insufficient homogenisation of starting powders. The presence of chemically inhomogeneous system in Ce or Zr doped BaTiO3 samples may be the cause of the decrease of dielectric constant. Table I: The grain size of Ce and Zr doped BT Sample No.

Composition of the sample

Average grain size of Zr doped BT(m)

BaTiO3 Ba0.9Ce0.1TiO3 Ba0.8Ce0.2TiO3

Average grain size of Ce doped BT(m) 1.024 1.69 1.31

1 2 3

0.0 0.1 0.2

BaTiO3 Ba(Ti0.9Zr0.1)03 Ba(Ti0.8Zr0.2)03

1.024 4.158 2.113

0.3

Ba0.7Ce0.3TiO3

1.42

Ba(Ti0.7Zr0.3)03\

1.538

Table 1 shows the grain size of Ce and Zr doped BT. The grain size of Ce doped BT is smaller than the grain size of Zr doped BT. Figure 2(a): EDS spectra of pure BaTiO3 Figure 2(b): EDS spectra of Ce-doped BaTiO3

Figure 2(c): EDS spectra of Zr-dopped BaTiO3

Fig(2a), Fig(2b), Fig(2c) shows the EDS spectra of pure and Ce and Z doped BaTiO 3respectively. EDS spectra confirms the presence of elemental compositions. grain size of ce doped BT

grain size in

4.5 4

grain size zr doped BT

3.5 3 2.5 2 1.5 1 0.5 0

0.1 0.2 0.3 concentration of Zr and Ce in BT

0.4

Figure 3: Change of grain size with concentration of Ce and Zr doping Fig.3 Shows the Change of grain size with percentage of doping element. The change in grain size with percentage composition of Ce and Zr shows the similar trend. As the concentration increases grain size increases upto x=0.1 and then decreases in both cases of doping.

Page 17

S.N. Rahman et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 15-19

IV. Dielectric measurements The effect of microstructure of dielectric properties of pure and doped BaTiO 3 at high frequency was studied experimentally. The observed microstructural features together with the type of additives have a direct influence on the dielectric properties of doped BT. The dielectric constant and loss factor of the samples were measured as a function of frequency[10], [11]. The effect of Zr and Ce in pure BT are shown in Fig.4 and Fig.5. Fig.4 shows that the presence of doping element such as Ce and Zr increases dielectric constant compared to pure BaTiO3. The higher value of dielectric constant and low loss in low frequency is found for the sample Ce doped BT than Zr doped BT. In Ce doped BT, small addition of CeO2 (x=0.1) increases grain size as well as dielectric constant compared to pure BT. It is observed that tan as well as dielectric constant decreases with the increase of Ce content (>0.1). Similar behavior is observed in the case of Zr doped BT. Figure 4: Frequency Vs Dielectric Constant of Figure 5: Frequency Vs. tan of Ce and Zr Ce and Zr doped BaTiO3 doped BaTiO3 1600 1400 1200 1000

pureBT 0.1CeBT 0.2CeBT

0.25

0.3CeBt 0.1ZrBT 0.2ZrBT

0.2 tand

Dielectric constant

0.3

pureBT 0.1CeBT 0.2CeBT 0.3CeBT 0.1ZrBT 0.2ZrBT 0.3ZrBT

800

0.3ZrBT

0.15

600

0.1

400 0.05 200 0

0 0

5000

10000 15000 20000 25000 30000 Frequency in KHz

5000

10000

15000

20000

25000

30000

Frequency in KHz

Ac field dependence on loss factor is illustrated for samples doped with Zr and Ce in Fig.5. In Ce doped BT samples the relaxation occurs at 1MHz for all the samples except in case of x=0.3 and Zr doped BT sample shows no significant relaxation in our present experiment. VI. Frequency response resistivity The resistivity at room temperature in the frequency range 75KHz-30MHz shows the resistivity  which falls to lower value than at higher frequency region. With the increase of CeO2 and ZrO2 , the resistivity of the samples show same characteristics up to x=0.2 for all samples but the value of the resistivity as well as dielectric constant increase for further addition of Zr and Ce in BT in the present experiment. Resistivity of the samples also depends on the grain size. It is evident that Ce doped BaTiO 3 shows high resistivity than Zr doped BT as shown in Fig.6 and Fig.7 which corresponds to high dielectric constant subsequently. Figure 6: Value of x in BaxCe1-xTiO3 Figure 7: Value of x in BaZr1-xTiO3x 1000000

1000

resistivity in ohm-cm

100000 10000

5MHZ 10 MHz 15 MHz 20 MHz 25 MHz 30 MHZ

Resistivity in Ohm-cm

5 MHZ 10 MHZ 15 MHZ 20 MHZ 25 MHZ 30 MHZ

100

1000 100

0.1

0.2 0.3 0.4 concentration of cerium

0.5

0.05

0.1

0.15

0.2

0.25

0.3

0.35

concentration of Zr

VII. Conclusions Comparative investigation of microstructure and dielectric properties of ceramics doped with Ce and Zr have been studied. Different microstructure regions were observed in the samples. The highest value of permittivity

Page 18

S.N. Rahman et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 15-19

was observed in the case of x=0.1 Ce doped BaTiO3 at frequency 75Hz which is characterized by fairly uniform microstructure with a homogenous distribution of additives. Dielectric constant is influenced not only by doping elements but also by concentration of doping element and grain size. A high dielectric constant (1485) was attained for the doping element CeO2 for x=0.1 at frequency 75KHz with low dissipation factor. The relaxation peaks are mainly due to the diffusion of oxygen vacancies induced by the acceptor impurities in an electrical field induced by variation of polarization of the material under electrical excitation. References [1]

M. Khan, “Preparation of small-grain and large grain ceramics from Nb doped BaTiO3”J. Advance chemical society, vol. 54(1), 1971, pp. 452-454. DOI: 10.1111/j.1151-2916.1971.tb12383.x

[2]

M. N. Rahman, R. Manalert, “Grain Boundary mobility of BaTiO3 doped with alio-valent cations”, J. Eur. Ceram. Soc. Vol. 18( 8), August 1998, pp. 1063-1071. doi.org/10.1016/S0955-2219(97) 00215-X.

[3]

S. K. Chiang, W. E. Lee, D. W. Ready, “Core shell structure in doped BaTiO3”, Am.Ceram. Bulletin 66(8), 1987, pp 1230.

[4]

D. Hennings, G. Rosenstein. Temperature stable dielectric based on chemically inhogeneous BaTiO 3, J. Am. Ceram. Soc. 67, 1984, pp 249-254.

[5]

R.T.Armstrong, R.C. Buchanan “Influence of core-shell grains on the internal stress state and permittivity response of Zirconiamodified Barium Titanate” J. Am. Ceram. Soc. vol.73[5], May 1990, pp 1268-1273. doi:10.1111/j.1151-2916.1990.tb05190x

[6]

A.Yamaji, Y Enomoto, K.Kinoshita and T. Murakami, “Preparation, Characterization and properties of Dy-doped small grained BaTiO3 ceramics”, J. Am. Ceram. Soc. vol. 60[3-4], March 1977, pp 97-101, doi:10.1111/j.1151-2916.1977tb15479.x

[7]

H.Y.Lu, J. S. Bow and W. H. Deng, “Core shell structures in ZrO 2-modified BaTiO3 ceramic”, J. Am. Ceram. Soc.vol. 73[12], Dec 1990, pp 3562-3568,doi: 10.1111/j.1151-2916.1990.tb04258.x

[8]

J. H. Hwang and Y. H. Han “Electric properties of Cerium-doped BaTiO3” J. Am. Ceram. Soc.vol. 84[8], Aug 2001, pp 1750-1754, doi: 10.1111/j.1151-2916.2001tb00910.x

[9]

D. Makovec and D. Kolar, “Internal oxidation of Ce3+-BaTiO3 Solid Solutions”, J. Am. Ceram. Soc.vol. 80[1], Jan 1997, pp 51-52, doi:10.1111/j.1151-2916.1997tb02789.x.

[10] S. N. Rahman, N. Khatun and Tofazzol Hossain, A.H khan “Dielectric studies of Cerium doped BaTiO 3 at different temperature. Jour. of Bang. Accademy of Sciences, vol. 32, No.1, 2008, pp 79-85. [11] Shamima Choudhury, Shurayya Akter, M.J. Rahman, A.H.Bhuiyan, S. N. Rahman, N. Khatun and M.T. Hossain “Structural, dielectric and electrical properties of Zirconium doped BaTiO3 perovskite” Jour. of Bang. Accademy of Sciences, vol. 32, No.2, 2008, pp 221229.

Acknowledgments The authors gratefully acknowledge the Director, BCSIR Laboratories, Dhaka of Bangladesh council of science and Industrial Research (BCSIR) for doing this research work.

Page 19

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Structural and Electrical Characterization of Ni-Zn Ferrites Md. Shahjahan1*, N. A. Ahmed1, S. N. Rahman1, S. Islam1, N. Khatun1 1 Industrial Physics Division, Bangladesh Council of Scientific & Industrial Research (BCSIR), Dhaka-1205, BANGLADESH Abstract: The effect of Zn substitution on the DC electrical resistivity, AC electrical conductivity and microstructure of V2O5 doped Zn(1-x)NixFe2O4 with x=0.0, 0.1, 0.2, 0.3, 0.4 (5 samples) were prepared by conventional ceramic technique has been investigated. The structural characterization of nickel-zinc ferrites were done by X-ray diffraction technique. Micro-structural and morphological studies were carried out by scanning electron microscope technique. The lattice constant determined from XRD data is in the reported range (8.4086 A.U.) DC electrical resistivity exhibits excellent semiconducting behavior. The result shows that AC conductivity increase with the increases in frequency. The micrographs of the samples shows that the average grain size increases with the decreases of Zn content while the grain size decrease with decreasing of Ni content. The variation of DC electrical resistivity with temperature is explained in this work. Keywords: DC resistivity; Electrical; Grain size; Microstructure; Transition temperature; Structural; I. Introduction In 1928 Forestier [1] prepared ferrites by heat treatment. Neel [2] in 1948 developed the model of ferrimagnetisms as a distinct class of magnetism. In 1985 A. B Naik and J. I Power [5] studied the dependence of resistivity and activation energy of Ni-Zn ferrites on sintering temperature and porosity. Ferrites are considered as soft magnetic materials [4]. The most important types of ferrites are manganese–zinc (Mn–Zn) and nickel–zinc (Ni–Zn) ferrites [6-7]. The resistivity of ferrites varies from 102 to 1010 ohm-cm which is up to 15 orders of magnitudes higher than that of iron [18]. The utility, variety and versatility make these materials highly demandable for high frequency application such as microwave devices, permanent magnets, electrical and component [8]. Ferrites are highly important electric materials widely used in electronic industries. Ferrites have most promising characteristics of excellent magnetic and electrical performance, high quality, low price and large number of controllable parameters etc [9]. Also transformer cores, rod antennas, radio frequency coils, multilayer chip inductors [10], wave absorbers, and converters. Recently, they were used as radar-absorbing materials. [7] At low frequencies, this interaction is small, and the eddy current losses are negligible. However, at high frequencies, the interaction causes unique phenomena, leading to application in microwave ferrite devices [11], such as telecommunication in cellular telephones and reception/transmission antennas [12], frequency-tunable oscillators and filters, isolators, circulators, and phase shifters [13]. Other specialized application of Ni–Zn ferrites are in the magnetic cores of read/write heads for high-speed digital tape or disk recording [14]. Ni–Zn ferrites have been intensely studied because of their remarkable high-frequency operation (1–100 GHz) as well as because they exhibit high chemical stability and high permeability in the radio frequency region [15]. In the present work our sample was V2O5 doped Ni-Zn ferrites by conventional ceramic technique. Our present investigation gives us an idea of the electrical properties of locally prepared materials. It has been shown that small amounts of Vanadium pentaoxide (V2O5) tend to remain in the grain boundary region, acting as liquid phase sintering aids [16]. II. Experimental Procedure All the reagents used for the synthesis of nickel-zinc ferrites were analytical grade and used as received without further purification. Zinc Oxide (ZnO), Ferosoferic Oxide (Fe2O3), Nickel Oxide (NiO) were used for the preparation of Zn(1-x)NixFe2O4 samples (where x=0.0, 0.1, 0.2, 0.3, and 0.4) and the raw oxide were collected from local market. The samples were prepared by conventional ceramic technique where ZnO and Fe2O3 were used as the parent materials and NiO were used as an additive. Samples were mixed with a mortar and pestle for fine particle size of the powders. Then the mixed was transferred into a porcelain dished were inserted into a central constant temperature controlled up to 700c. The dishes were kept at this temperature for 3 hours and then switched “Off” the furnace. After 18 hours the dishes were taken out of the furnace. The materials become lightly red color. The pre-sintering materials again mixed with mortar and pestle formed into uniform powders. Small quantity of pre-sintered powder was mixed well with some drops of poly-vinyl alcohol of 0.1% as a binder. V2O5 was used as additive in 0.1%, mole percentage. For the sintering requirements [17] we sintered the samples at 1200c for 3 hours in a crucible and then the furnace was switched “off” for slow cooling after 18

Page 20

shahjahan et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 20-25

hours the crucible was taken out of the furnace. The rough sides of the tablet were then polished on a fine grade emery paper. The samples were then cleaned to avoid any unwanted dust or impure particles. After that the samples were pasted with air drying type silver paste for electrical contacts. Sides of the sample were taken free for protecting short circuit. The tablet was then kept in an oven for twenty four hours for complete drying. The structural characterizations were carried out by the X-ray diffraction. XRD data were taken at a room temperature using Cu-Kα ( =1.5406 Å). A scanning electron microscope (SEM) was employed for the observation of the surface morphology and an estimate of grain sizes with increasing NiO content. These pellets were used for the measurement of the temperature dependent resistivity. A two-probe method was used for the measurement of the resistance and capacitance of the samples. The heat treatment was performed in a programmable furnace. III. Results and discussion A. Structural properties The Crystalline phases were identified using a Bruker X-ray powder diffractometer using  (Cu-Kα) =1.5406 Å. The x-ray diffraction (XRD) patterns of Ni-Zn ferrites were collected at room temperature with a step size of 0.02 2 and a counting time of 10s. The determination of the lattice constant and other structural parameters of the spin phases was made from the X-ray diffraction patterns. The structural parameters and atomic positions for the spinel phase were taken from the literature. Figure 1 shows the powder X-ray diffraction pattern (XRD) of nickel-zinc ferrites. All the peak belongs to the cubic spinel structure and analysis of XRD patterns prove the formation of single phase samples. The lattice constant “a” was calculated using the formula

a  d hkl h 2  k 2  l 2 Where h,k,l are the Miller indices and dhkl is the inter planer spacing. The lattice constant obtained from XRD data is in reported range (8.4086 A.U). The effects of Zn substitution V2O5 addition to Ni-Zn ferrites compared in the SEM micrographs shown in figure 2. The micro structure of NiZn ferrites exhibits homogeneous grain distributions. The grain size and transition temperature (T c) increase with the decreases of the Zn content while the grain size and Transition temperature (T c) decrease with decreasing Ni content in Ni-Zn ferrites as shown in figure 3. Table I. Different parameters of Ni-Zn ferrites Sample Composition ZnFe2O4 Zn0.9Ni0.1 Fe2O4 Zn0.8Ni0.2 Fe2O4 Zn0.7Ni0.3 Fe2O4 Zn0.6Ni0.4 Fe2O4

Average grain size, D(m) 4.3610 3.0007 3.6580 5.1650 5.5660

Resistivity,  in -m at room temperature. 1.960102 1.690104 2.090104 1.210103 6.895103

Transition temperature Tc(c) 215 150 165 295 310

Resistivity  in -m at Tc 1.978102 2.369104 3.598104 1.650104 2.363105

Activation energy Ep in eV at Tc 3.840010-4 1.23110-2 2.050910-2 1.555010-2 1.776010-2

Figure 1. X-ray diffraction pattern for Ni-Zn ferrites. (a) x=0.0, (b) x=0.1, (c) x=0.2, (d) x=0.3, (e) x=0.4

Page 21

shahjahan et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 20-25

Figure 2. Scanning electron micrographs taken on surface at 10m scale of Zn(1-x)NixFe2O4 samples sintered at 1200c during 18 h with (a) x=0.0, (b) x=0.1, (c) x=0.2, (d) x=0.3, (e) x=0.4 a

Page 22

shahjahan et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 20-25

350

300 250

200 3 150 2

Average grain size, D(mm)

100

Transition temperature Tc(°c)

Transition temperature Tc(c)

Average grain size, D µ-m

0 0

0.2

0.4

0.6

Value of x in Zn(1-X)Nix Fe2O4

Figure 3. Change of transition temperature (Tc) and grain size (D) due to Ni doping in Zn(1-X)Nix Fe2O4 B. Electrical properties B1. Effect of DC electrical resistivity The effect of appropriate V2O5 on the electrical behavior of doped Ni-Zn ferrites can be analyzed through resistivity Vs temperature curves as shown in figure 4. The DC resistivity was measured as a function of temperature and showed that the DC resistivity is almost same up to definite temperature in the case of each sample and then it falls into ups and downs. It continues up to knocking at the Curie temperature Tc. After Tc the DC resistivity decreases gradually with the increase of temperature. The resistivity of the samples ZnFe 2O4 at this transition temperature was found to be 1.978×10 2 -m. The resistivity of samples V2O5 doped Zn(1x)NixFe2O4 show the same characteristics, but the value of resistivity decrease with an increase of the doping concentration which is represented in table-I. The activation energy (Ep) increases with the decrease of Zn content. It also notes that Ni concentration shifts the Curie point of Ni-Zn ferrites to higher temperature. The

Page 23

shahjahan et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 20-25

sharp decrease in resistivity above the Curie temperature Tc can be explained on the basis of clusters of Zn 2+ ions [17].

100000

x=0.0

Resistivity in -m

x=0.1 10000

x=0.2 x=0.3

1000

x=0.4

100

10 0

Temperature in (c)

Figure 4. Resistivity in -m Vs temperature in c for samples B2. Effect of AC electrical resistivity The effect of different composition for Ni-Zn ferrites, AC conductivity as a function of frequency is shown in figure 5. The values of AC conductivity are depending on frequency. The polarization behavior of ferrites is the electronic exchange [19] between Fe2+Fe3+ [3] is proposed under this investigation of the frequency dependence of the AC conductivity. 2

x=0.0

a.c conductivity×10-3 mho/m

x=0.1 x=0.2

1.5

x=0.3 x=0.4 1

0.5

0 0

500

1000

1500

2000

2500

Frequency in kHz

Figure 5. AC conductivity Vs Frequency of the samples Table II. AC electrical properties measurements for five Ni-Zn ferrites samples Sample No

X=0.0 X=0.1 X=0.2 X=0.3 X=0.4

Conductivity (σa.c) in mho/m at frequency 75(KHz) 0.32710-3 0.09710-3 0.09510-3 0.08210-3 0.04310-3

2(MHz) 1.41210-3 0.56110-3 0.69910-3 0.33810-3 0.61310-3

Page 24

shahjahan et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 20-25

IV. Conclusion The DC electrical resistivity and micro structural properties of Ni-Zn ferrites were influenced significantly by small additions of V2O5. It was found that with an increase of Ni in Zn(1-x)NixFe2O4 (where x=0.0, 0.1, 0.2, 0.3, 0.4) the grain size of the samples increases. Samples 5 (x=0.4) exhibited the highest resistivity among the all samples. The resistivity increases with temperature up to a maximum at a temperature which is termed as the Curie temperature or the transition temperature (Tc) and then the resistivity decreases. The Tc rises with an increase of the value of x in Zn(1-x)NixFe2O4 as well as the grain size. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

[12] [13] [14]

[15] [16] [17] [18] [19]

Y. Kato and T. Takei, “Studies on Composition, Chemical Properties and Magnetization of Zinc Ferrite,” J. Japan Mining Society, vol. 539, 1930, pp. 244-255. L. Neel, Ann. Phys. 3 (1948) 137 P.Venugopal Reddy and T. Schsagiri Rao, “Dielectric behavior of mixed Li-Ni ferrites at low frequencies,” J. less-Common Met, vol.86, 1982, pp.255-261 R. Satyanarayana, Ph.D.Thesis, Osmania University, 1983 A. B. Naik and S. A. Patil, J. I. Powar, “X-ray and magnetization studies on Li-Cu mixed ferrites,” J. Materials Science Letters, vol.7, October 01, 1988, pp. 1034-1036 R. Valenzuela, Magnetic ceramics, Cambridge University Press, Cambridge, 1994. G. Herrera, M. M. Pe´rez-Moreno, “Microstructure dependence of the magnetic properties of sintered Ni–Zn ferrites by solidstate reaction doped with V2O3,” J. Mater Sci.vol.47, 2012, pp.1758–176, doi: 10.1007/s10853-011-5956-z B.V. Bhise, A. K. Ghatage, B. M. Kulkarni, S. D. Lotke, S. A. Patil, “Conduction in Mn substituted Ni-Zn ferrites,” Bulletin of Materials Science, vol.19, Issue 3, June 1996, pp.527-531. Encyclopedia of chemical technology 8, pp 88a. T. Nakamura, “Low-temperature sintering of Ni-Zn-Cu ferrite and its permeability spectra,” J. Magn. Magn. Mater, vol. 168, Issue 3, April 2, 1997, pp. 285-291. Lei Zhang, Xueyan Liu, Xingjia Guo, Mingming Su, Tianci Xu, Xiaoyan Song, “Investigation on the degradation of brilliant green induced oxidation by NiFe2O4 under microwave irradiation,” College of Chemistry, Liaoning University, Shenyang 110036, People's Republic of China Chemical Engineering Journal - CHEM ENG,vol.173, Issue 3, Jan01, 2011,pp.737-742, doi:10.1016/j.cej.2011.08.041 UR. Lima , MC. Nasar, RS. Nasar, MC. Rezende, V. Arau´jo, “Ni–Zn nanoferrite for radar-absorbing material,” J. Magn. Magn. Mater, vol. 320, Issue 10, May 2008, pp. 1657-1728, doi10.1109/22.17452. M. Buswell, “Modeling ferrimagnetic resonators, Microwave Theory and Techniques,” IEEE Transactions on Hewlett-Packard Co., Santa Rosa, CA, USA, vol. 37, Issue 5, May 1989, pp.860-887, doi.10.1109/22.17452. TT. Srinivasan, P. Ravindranathan, LE. Cross, R. Roy, RE. Newnham, SG. Sankar, KC. Patil KC, “Studies on high‐density nickel zinc ferrite and its magnetic properties using novel hydrazine precursors,” J. Applied Physics, vol.63, Issue 8, April 1988, pp. 3789-3791, doi. 10.1063/1.340615 CW. Chen,“Magnetism and Metallurgy of Soft Magnetic Materials,” North Holland, Amsterdam, Dover Books on Physics, Paperback, 1977 SH.Chen, SC. Chang, CY. Tsay, KS. Liu, IN. Lin, “Improvement on magnetic power loss of MnZn-ferrite materials by V2O5 and Nb2O5 co-doping” J. European Ceramic Society, Vol. 21, Issues 10–11, 2001, pp. 1931–1935 S. Hafner, Schweiz, Min, petrogr., Mitt, 40, 1960, 207. D. Ravinder and K. Vijay Kumer, “Dielectric behavior of erbium substituted Mn-Zn ferrites,” Bull mater. Science, Vol.24, Issue 5, October 2001, pp. 505-509. K. Radha K and D. Ravinder, “Frequency and composition dependence of dielectric behavior of mixed Li-Cd ferrites” Ind. J. Pure and appl. Physics. Vol. 33, 1995, pp.74-77.

Acknowledgments We would like to express our grateful thanks and gratitude to the authority of BCSIR for providing us the opportunity and necessary permission to carry out this research work.

Page 25

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Comparative Study of Web Page Ranking Algorithms Atul Kumar Srivastava1, Mitali Srivastava2, Rakhi Garg3, P. K. Mishra4 Department of Computer Science, Faculty of Science1, 2, 4, Mahila Maha Vidyalaya3 Banaras Hindu University, Varanasi, (U.P.), INDIA1, 2, 3, 4 Abstract: With the exponential growth of information on web, getting relevant information regarding user query through search engines is a tedious job today. Several search engines use link analysis algorithms to rank the web pages according to the need. But these algorithms are still lacking with efficiency, scalability and relevancy issues. This paper put forward survey of various improved ranking algorithms and their pros and cons. Further, we have included comparative study of various ranking algorithms mainly PageRank and HITS based on computation environments like Sequential, Parallel. This will help scientist, researchers, and academicians working in this area to understand the existing algorithms and develop one which is need of today’s environment. Keywords: Web mining, Web structure mining, PageRank algorithm, HITS algorithm, Improved PageRank algorithms, Comparison of various page rank algorithms. I. Introduction World Wide Web (WWW) has become very popular and interactive medium to broadcast information. With the exponential growth of the Web, there is a huge amount of data and information available in web, so that accessing information from web with accuracy and speed is a big challenge for both people and software. Many problems with web exist like personalization of web pages, finding useful information, creating the knowledge based on extracted information and Learning about consumers and users. These problems arise due to multiplicity of data (i.e. structured, semi structured and unstructured) present in a web, containing many redundant information, due to dynamic nature of a web etc. [2, 11] Web mining is emerged as a broad research area to solve these issues in last few decades. The web mining field is a converging field with Database, Information Retrieval and Artificial Intelligence [1]. Web mining is the application of data mining techniques to discover and extract the useful and previously unknown information from the web data. Web mining can be classified into three main categories according to the type of data to be mined [3, 5]:

Figure 1: Categorization of web mining [3,5]

It is difficult to retrieve relevant and useful information from web because of existence of data in large amount and of its heterogeneous nature. As a result of huge and heterogeneous data available on WWW user get thousands or millions of web pages related to their query through search engine. Web users do not have much time and patience to go through all returned pages to find the relevant information that are of their interest and use [2, 3]. Web structure mining uses social network analysis and various link analysis algorithms to help web users to get information of their use on time. Link analysis or web page ranking algorithms rank web pages based on authority, popularity and prestige of web pages. The role of link analysis algorithm is to select the web pages that are most likely be able to satisfy the user’s need and provide them higher ranks [13]. There are two main Link analyses or ranking algorithms: PageRank (used by Google) and HITS.Page rank algorithms are used to rank web pages according to the relevancy. Link analysis algorithms try to keep the desired result within the top few pages, otherwise, the web search engine could be considered as unserviceable. There are following issues with the basic link analysis algorithms [3, 7, 14]: a) Uniformly sampling of web pages b) Discovering duplicate hosts c) To model in web graph

Page 26

Atul K. Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 26-32

d) To avoid web spamming e) To discover the content quality of web page In this paper, we have reviewed page rank algorithms and focus on their pros and cons. Section II includes web structure mining in details. In section III, we reviewed two popular link analysis algorithms (sometimes referred as page ranking algorithm): PageRank and HITS algorithm, and their improvements over the decades. Section IV shows comparative study of ranking algorithms in environment like Sequential, Parallel and distributed. Finally, Section V concludes the Paper. II. Web Structure Mining Web Structure mining is a process which discovers the structural summary about the hyperlinks and web pages. Web structure mining could be divided basically into two categories: Document structure analysis and Hyperlink analysis [9].Document Structure analysis provides structural information of web pages i.e. how contents are organized in HTML and XML tags. On the other hand Hyperlinks analysis provides how the web pages are connected with each other. Link information is used to combine with content of web to measure quality of information. It is also useful to find community of web pages according to userâ&#x20AC;&#x2122;s common interest. Many researchers have summarized the following some basic goals for determining the quality of individual web pages and group of web pages [5, 9]: a) To count the local links in Web tuples (rows in a web table) in web table (representation of web pages). Local links are those links which connects different web documents which are related to same server. This also specifies the completeness of a particular web site. b) To count the links in web tuples which are global in nature and the links spans different web sites. This describes visibility of web page and ability to relate same or other related documents across the web. c) To measure the frequency of identical web tuples that appears in the indexed web pages or among the Web tables. Link analysis provides an effective tool to calculate the importance of web pages on any particular topics. To search any web page from the search engines involves two main steps: The first step extracts the relevant pages according to userâ&#x20AC;&#x2122;s query while the second step rank pages based on their quality [8, 10]. There are many different methods proposed to rank the web pages using hyperlinks. Following section include the brief discussion about these methods. III. Link Analysis Algorithms Every web search engine uses its own ranking algorithm to put resulted web pages in decreasing order of relevance so that appropriate result would be on first page. Ranking algorithm uses hyperlink analysis of web graph to rank the web pages. There are many different method proposed to identify the relevancy of web pages, some users use random walk on web graph while others uses web graph structure. There are two main ranking algorithms: PageRank (uses by Google) and HITS. Both of them rank relevancy score of individual web page. Some common link analysis algorithms are discussed below:A. In-degree Algorithm: This is the oldest algorithm which ranks the web pages according to the popularity of web pages [13]. Popularity of a particular web page is measured by the number of incoming links; if any page has many incoming links then the page is more popular than less number of incoming links. Mathematically, it can be shown by the following equation [13]: (1) Where denotes the in-degree of that particular page and denotes the rank of that page. This algorithm was previously used by AltaVista, HotBot and several other web search engines. B. PageRank Algorithm In-degree algorithm ranks the web pages according to the number of incoming links. Brin & Page [4], who is the founders of Google search engine, observed that the number of incoming links is not the only criteria to determine rank. Further, they extended this algorithm and observed that all incoming should not be given equal importance. A web page should assigned higher rank if it is referenced by many high ranked web pages. By using this concept they proposed the simplified algorithm called as PageRank algorithm which computes rank of web pages iteratively. Mathematical equation for this algorithm is stated as [4, 15]:

Where and are the web pages, denotes PageRank values of web page , is the set of web pages point to page , indicates the pages that link to page u, and c is normalization constant (c<1). Table 1 summarizes few advantage and disadvantage of PageRank algorithm [4, 6, 27, 30]:

Page 27

Atul K. Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 26-32

Table 1: Some observations of PageRank Algorithm 



Advantage Capable to fight with spam web pages: In PageRank algorithm a page assigned higher rank if the pages pointing to it have high rank. Due to this it is not easy task for a particular web page owner to add several incoming links to the pages which pointing to his/her page, i.e. it is not easy task to influence PageRank. Global rank score computation: PageRank algorithm computes the rank score of web pages globally i.e. it computes of the rank web page of whole web graph unlike to HITS. Query independent algorithm: It computes the rank of web pages of whole web graph globally and stored off-line rather than query time. At query time search engines just match the query keyword with some other strategies to rank the web pages and sorted them according to the rank value of web pages. Efficient and fast computation: Due to global computation and query independent nature it is very fast and efficient algorithm. These two advantages play very important role in the success of “Google” search engines.



 



Disadvantage Rank Sink: Let consider a situation where a sub graph of web graph in which two web pages point to each other and do not point to any other page, and other web pages point to any one of these two web pages. So during the iterative computation of rank this loop accumulates the rank score and do not distribute rank score to any other web page. Dangling Node: Dangling nodes are the nodes in web graph that have no-out edges. So they are not able to distribute their ranks. It affect the model when the number of dangling nodes is too large. Recency Search: The important factor in PageRank algorithm is the number of incoming link to any page. The old pages in web have more incoming link than newly crawled web pages so old pages get higher rank value than new ones. But relevant information may be in newer pages. Topic drift: PageRank is query independent algorithm. So it is unable to predict whether a hyperlink is related to user’s query or subject of web pages. In result of this it may suffers with topic drift problem. Link spamming (Page Cheating): Some commercials web sites use optimization of search engines try to get position of first page in search engines results. It is obvious that PageRank only consider number of incoming links not the content of pages.

C. HITS Algorithm According to Brin and Page, PageRank algorithm is one level wait propagation scheme. Later Kleinberg proposed a different query dependent ranking algorithm which is based on two level weight propagation schemes [12]. In HITS algorithm rank of every web page is determined by two score rather than one in PageRank: First score is authoritative score which measures the quality of page by containing relevant or valuable information regarding query, and Second score is hub score which measures the quality of page by containing link of authority pages. Clearly, a good authority page is a page which pointed by many good hub pages and a good hub page is a page which points to many good authority pages. We can bipartite web graph according to hub and authority pages by making two parts of each page, where one set contains hub pages and other contains authority pages. Computation of HITS: Computation of hits algorithm mainly done in two steps: In first steps collect the web pages on which hits algorithm applied and second steps perform hits computation in those web pagesStep 1: HITS algorithm is applied to a subset of web graph. The subset of web pages is the collection of relevant pages depending on the query. The algorithm starts with a root set of the top t pages from the result list of query q, by some content-depend rank algorithms ( from a text based search engines ex- Alta-vista or HotBot). From the root set a neighboured set is obtained by [12, 13]:

Where denotes the neighbored set on which authority and hub score would be computed. Step-2 Computing Authorities and Hubs: To calculate the hub and authority score for page Kleinberg defined that authority score of a page is the sum of all hub score of pages that point to it and hub score of a page is the sum of all authority score of pages that is pointed by it. There is mutual reinforcement relationship between hubs and authorities, by make use of this relationship updates the hub and authority score of each page iteratively [12]. , Where h, a represents n-dimensional vector of hub and authority score of n pages. Issues with HITS algorithm: There are several issues with HITS algorithm [31, 33, 34] a) The main advantage of HITS algorithm is its query dependent nature i.e. it computes the rank of web pages based on user’s query. But due to query time evaluation this algorithm is not feasible with respect to time because search engines contain billion to trillion of query per day. b) This algorithm is not capable to fight with spam web pages as it computes the rank of web pages based on hub and authority score. c) This algorithm also suffers from topic drift problem. When it collect the web pages for root set it does not uses its own algorithm, so maybe it collects the web pages which are not relevant to query topic.

Page 28

Atul K. Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 26-32

D. Summary of Improved PageRank algorithms Several researchers have proposed improved PageRank algorithms to overcome issues as stated in table 1 with basic PageRank algorithms. Categorizations of improved PageRank algorithm based on their issues are summarized in table-2: Table 2 - Summarization of modified ranking algorithms based on handling issues Algorithm designed by Authors Sergey Brin and Lawrence Page [4]

Handling issue

Ideology for the method

Rank Sink Problem

Uses Random surfer model to get out of the loop by using damping factor

Sergey Brin and Lawrence Page [4]

Dangling Link

Remove the entire dangling link until all PageRank are calculated and after calculation simply add them.

Sung Jin Kim and Sang Ho Lee [21]

Dangling Link

ShiguangJu, Zheng Wang and Xia Lv [28]

Recency Search

Define a new matrix A in which all dangling column have value and then compute the PageRank (A*) value by using this matrix Proposed a new method which combines the last modification time of web pages with the in-link and out-link weight of concerned web pages.

Wenpu Xing and Ali Ghorbani [16]

Recency Search

Jing Wan and Si-XueBai [32]

Recency Search

Zhou Cailan and Chen Kai [27]

Topic Drift Problem

Xiaoyun Chen, BaojunGao and Ping Wen [22] Chia-Chen Yen and JihShih Hsu [29]

Topic Drift Problem

AndriMirzal and Masahi Furukawa [37]

Topic Drift problem

BunditManask asemsak and ArmonRungsa wang [18] Yizhou Lu, Benyu Zhang, Wensi Xi et.al. [35]

Computation time and Resource

Atish, Anisur and Eli [36]

Scalability

Mathematical formulation of the method

Use stochastic matrix to find the dangling links

Where

denotes PageRank matrix

and is the In-link weight and out link weight of link (q, p) and is links decay weigh

Link Spam Problem

Efficient Computation

Proposed a new method which considers the indegree and out-degree of web pages and distributes the rank based on the importance of pages.

Combines the time activity model (based on web site, user interest, content of web pages and web site developer) with the traditional PageRank algorithm. Proposed a method which maintains user search log based on random query and click log file, which contains click time and information about URL’s. This method add an attribute with PageRank algorithm i.e. Click weight. Proposed a method which distributed the rank of web page to its outgoing page is based on page similarity, and latent semantic model is used to determine the similarity between web pages. In the proposed method the PageRank score of web pages does not distributed equally by the number of out-degree, but only relevant web pages can share the score according to relevance degree. Proposed a link-viewer tool to observe the topic drift problem, Solve these issues in two steps: in first step find out the most relevant web pages for the root set by projection method. In next step extract out the web pages which do not have many out links from the root set. This method speedup the PageRank computation by partitioning the URL index files by source URL into n files of same size called : 0 n and each files is executed on separate processor. Proposed method is based on two attributes of web graph i.e. “Power Law Distribution” and “Hierarchy Structure”. First it finds the low scorer pages then compute the global PageRank by combining these low scorer pages. Proposed a scalable distributed model which used Monte-Carlo method to compute PageRank of web graph.

Where and

is the weight of incoming link weight of outgoing link

Where TR denotes the time-activity importance vector of page. If TR value is more that means the web page importance is more at that time.

Where denotes click weight and denotes weight of click time

Where

denotes

similarity between web pages.

where is the relevancy score between web page j to i. Uses Projection and Base-set Downsizing method to filter out the irrelevant pages from root set.

Where rank vector which computed parallel. Power law distribution is used to for finding the importance of web pages in web graph.

Implement random surfer model based distributed algorithm to compute PageRank of web graph.

Page 29

Atul K. Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 26-32 ArmonRungsa wang and BunditManask asemsak [19, 26]

Efficient Computation

Present the web graph in compact format so that after partitioning the graph into small parts it should be fit into the GPU’s memory devices called cluster’s. It takes advantage of GPU’s and gives better result than parallel PageRank.

Compute the PageRank value Parallel on GPU system by using OpenMPI and GCC.

Yuan Wang and David J, DeWitt [24]

Scalability

This method proposed a framework which computes PageRank in distributed manner i.e. every web server computes PageRank locally over its own data and then result from every server merged to generate an indexed to determine global ranking.

Compute local Page Rank vector in individual web server Where, represents the server m that contains pages and is the uniform column vector of dimension . Compute the server rank vector Where Gs represent the server link graph.

Shumingshi, Jin Yu, and Guang Wen Yang [38]

Scalability

In this paper distributed page rank proposed by using open system PageRank method. Indirect transmission is used to reduce the communication overhead between servers.

Compute Y in distributed manner. Where Y be the rank vector B is square matrix with and R ranks of all pages in the groups.

IV. Comparitive Study Table III, IV consecutively describes the comparative study of sequential algorithms and parallel Algorithms. These comparisons are based on parameter like method type, additional attribute, relevancy score, input, computation time, merit and issues. Table 3 - Comparison of Sequential Ranking Algorithms Parameter Methods HITS

PageRank (PR)

Weighted PageRank (WPR)

Latent Semantic Model(LSM)PageRank

Feedback of user click – PageRank [27]

Method Query dependent method which computes the importance of web page based on two score: Hub and Authority. Query independent method which sorted the result based on the relevancy score of web pages Modified basic PR by Computing the relevancy of web pages based on the weight of in-links and out-links. Modify PageRank by distributing the rank of web pages to outgoing link based on the having similar contents in pages Modification to PageRank by Combining PR method with the user feedback of search results.

TimeActivity based PageRank [32] Page Relevancy based PageRank [29]

Modify basic PR by Computing the rank value of web pages based on time activity curve. Combined basic PR with the user feedback of result by calculating relevancy score between web pages.

Power-Rank Algorithm [17]

Computes the rank score in two step: In first step it filters out low score web pages

Attribute used Hyperlink and content structure

Random surfer model to avoid rank sink issue Weight of incoming and outgoing links LSM is used to determine similarity between pages HTTP log file for the feedback of the search result TimeActivity important level model Relevancy score to distribute rank of pages to other pages Power law distribution

Input

Relevancy

Complexity

Merits

Issues

Incoming link + Outgoing link + Content of web pages Incoming link

< PR

<O(log N)

More relevance than PR based on content

> HITS

O(log N)

Less time to give output of search results than HITS.

Time complexity is more than PR to give output result, Topic Drift, link Spamming Problem etc. Topic Drift, Recency Search and Dangling node Problem.

Incoming link + Outgoing link

<PR but > HITS

<O(log N) but >HITS

More quality pages than PR

Lacking with Recency Search Problem

Incoming links + similarity score between pages PR + Click weight + Weight of click time PR + Time activity important degree Incoming links + relevancy score of web pages Incoming links

< PR

> PR

More relevant pages than PR by resolving Topic Drift problem

More complex than PR due to finding the similarity score between pages.

<PR and WPR

> PR

Resolve topic drift and Recency Search Problem

More computation time due to use of Web log Mining.

< PR

> PR

Solve Recency Search Problem

More complex than PR

< PR

> PR

Solve link spamming problem and convergence rate< PR

More complex to calculate rank value than PR.

= PR

<PR

Less computation time and filter out the

Possibility to filter out some new relevant pages

Page 30

Atul K. Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 26-32 and in next step it computes the global importance score of pages. WTPR [28]

duplicate pages

Associate basic PR with time attribute i.e. last modification time of web pages and hyperlinks

Log-file gives the modification time of page & link

In-links +Timestamp of node & links

< PR

> PR

Solve Recency search issue and suitable for dynamic changes of web

More complex due to dynamic nature and takes more time than PR

Table 4 - Comparison of Parallel and distributed ranking algorithms Parameter Algorithms

Method

Attribute used

Input

Relevancy

Time complexity

Merits

Issues

Yuan Wang David J.DeWitt [24]

Computes the rank score of pages on individual server then merge the overall rank of pages to get the result. Divide the web graph in several parts then apply PageRank method on low cost Parallel system

Distributed system to enable scalability

Intra server links, inter server links

Relevancy is same as PR

Less time than Sequential PR when dealing with large data.

Highly Scalable

Merging of rank value of web pages returned by individual server is too complex

MPICH v 1.2.5 for Parallel computation, and PC Cluster for storing Partitioned web graph GPU used for storage and to compute rank value on Meka cluster, and Open-MPI Monte Carlo method with distributed environment

Equally partitione d web graph into each PC cluster

More quality pages are returned than sequential PR

Time complexity will decrease when no. of processors increases till the threshold time

Low cost parallel system to compute large web graph.

Takes more computation time when number of cluster increases

Smaller size of web graph that fit in GPU cluster

Approximate ly Same as Sequential PR

Computation time is less than sequential PR

Fast computatio n with no constraint on the size of web graph

Takes more time on the copy of data between GPUs to CPU devices than other part of the task.

Same as PR

More quality pages than Sequential PR and HITS method

Takes to less computatio n time than other distributed algorithm

Used in large-scale, distributed network & resource constrained system where computation time is important.

BunditManas kasemsak and ArnonRungsa wang [18]

A. Rungsawang and B. Manaskasems ak [19]

Computes the PageRank value on Graphical Processing Unit (GPU) system.

Sarma, Molla, GopalPandura ngan, and EliUpfal [36]

Implement the fast random server based PageRank algorithm in distributed environment

for undirected graph,

for directed graph

V. Conclusion Web structure mining is one of the categorization of web mining which mines the hyperlink structure of web graph. Now days search engines are producing trillions of web pages based on user’s search query. Ranking algorithms plays a vital role to find the quality of web pages. So efficiency, relevancy and scalability of ranking algorithms are important issues. Various ranking algorithms like HITS, PageRank and their modifications were proposed. This paper has included study of PageRank algorithm and their modification based on disadvantages of basic PageRank algorithm. It also explores modified algorithms in various environments used by researchers like parallel distributed and their pros and cons. Sequential modified PageRank algorithms almost resolved issues (Rank Sink, Dangling node, topic drift etc.) with basic PageRank algorithm but they are still lacking with some other prospects like scalability, relevancy etc. In Future Parallel, Distributed, Grid environment could be used to solve Topic Drift, Recency search and scalability issues by applying on various modified PageRank algorithms. References [1] [2] [3] [4] [5]

R. Kosala, HendrikBlockeel (2000), “Web Mining Research: A Survey”, ACM SIGKDD Explorations, Vol.2 Issue 1, Page(s):1-15. S. Chakrabarti, B. E. Dom, D, Gibson, J. Kleinberg, R. Kumar, P. Raghavan, S. Rajagopalan, and A. Tomkins (1999), “Mining the Link Structure of the World Wide Web” IEEE Computer, Vol.32, Page(s):60-67. A. Mülle(2011), “Web Mining in Social Media: Use Cases, Business Value and Algorithmic” by Social Media-Verlag Publication. S. Brin, L. Page(1998), “The Anatomy of a Large-scale Hyper textual Web Search Engine” Proceedings of the Seventh International World Wide Web Conference, Page(s):107-117. Wang Jicheng, Huang Yuan, Wu Gangshan, Zhang Fuyan (1999), “Web mining: knowledge discovery on the Web”, IEEE International Conference, Vol.2 Page(s):137 – 141.

Page 31

Atul K. Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 26-32 [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38]

Taher H. Haveliwala (2003), “Topic-Sensitive PageRank: A Context-Sensitive Ranking Algorithm for Web Search” In IEEE Transactions on Knowledge and Data Engineering, Vol. 15, No4, Page(s):784-796. Monika R.Henzinger, R. Motwani and C. Silverstein (2002), “Challenges in web search engines” SIGIR Forum, 36(2): Page(s):11-22. Bing Liu(2011), “Web Data Mining, Exploring Hyperlinks, Contents, and Usage Data” Second Edition, Springer. P Desikan, J Srivastava, Vipin Kumar, and Pang-Ning Tan(2002), “Hyperlink Analysis: Techniques and Applications” Page(s):1-42. A.Broder (2002), “Web Searching Technology Overview”, advanced school and Workshop on Models and Algorithms for the World Wide Web. Qingyu Zhang, Richard S. Segall (2008),” Web Mining: A Survey of Current Research, Techniques, and Software” International Journal of Information Technology & Decision Making, Vol.7 Issue 4, Page(s):683-720. J.Kleinberg (1999), “Authoritative sources in a hyperlinked environment”, Journal of ACM, Vol.46 Issue 5, Page(s):604-632. A.Borodin, G.O.Roberts, J.S.Rosenthal, P.Tsaparas (2005), “Link analysis ranking: algorithms, theory, and experiments”, ACM Transactions on Internet Technology, Vol.5 Issue 1, Page(s):231-297. F.McSherry (2005), “A Uniform Approach to Accelerate PageRank Computation”, Proceedings of the 14th World Wide Web Conference, Page(s):575-582. PavelBerkhin (2005), “A survey on PageRank computing”, Internet Mathematics 2, Vol.1, Page(s):73–120. Wenpu Xing and Ali Ghorbani (2004), "Weighted PageRank Algorithm", Proceedings of the Second Annual Conference on Communication Networks and Services Research (CNSR'04), IEEE, Page(s):305-314. Lu, Y., Zhang, B., Xi, W., Chen, Z., Liu, Y., Lyu, M.R., Ma, W.Y (2004), “The PowerRank Web Link Analysis Algorithm” In: 13th WWW conference, Page(s):254–255. BunditManaskasemsak and ArnonRungsawang (2005), “An Efficient Partition-Based Parallel PageRank Algorithm” Proceedings of 11th International Conference on Parallel and Distributed Systems, IEEE, Vol.1, Page(s):257-263. A. Rungsawang and B. Manaskasemsak (2003), “Fast PageRank Computation using PC Cluster”, In Proceedings of the 10th European PVM/MPI User’s Group Meeting, Vol.2840, Page(s):152-159. K. Sankaralingam, S. Sethumadhavan, and J.C. Browne (2003),” Distributed PageRank for P2P Systems”, In Proceedings of the 12th IEEE International Symposium on High Performance Distributed Computing, Page:58. Sung Jin Kim and Sang Ho Lee (2002), “An Improved Computation of the PageRank Algorithm”. In Proceeding of the European Conference on Information Retrieval (ECIR), Page(s): 73-85. Xiaoyun Chen; BaojunGao; Ping Wen(2009), “An Improved PageRank Algorithm Based on Latent Semantic Model” In proceeding of: Information Engineering and Computer Science, Page(s):1-4. R Khare, D Cutting, K Sitaker, Rifkin (2004), “Nutch: A flexible and scalable open-source web search engine”. Oregon State University Yuan Wang David J. DeWitt (2004), “Computing PageRank in a Distributed Internet Search System” Proceedings of the 30th VLDB Conference, Toronto, Canada, Vol.30, Page(s):420-431.. Ali Mohammad ZarehBidoki, Nasser Yazdani (2007), “Distance Rank: An intelligent ranking algorithm for web pages”, Information Processing and Management in Elsevier, Vol. 44 Issue 2, Page(s):877-892. B. Manaskasemsak and A. Rungsawang (2004), “Parallel PageRank computation on a Gigabit PC cluster” In Proceedings of the 18th International Conference on Advanced Information Networking and Applications, Vol.1, Page(s):273-277. Z. Cailan, C. Kai,LiShasha (2011), “Improved PageRank Algorithm Based on Feedback of User Clicks”, in IEEE, Page(s):3949-3952. ShiguangJu, Zheng Wang, Xia Lv (2008), “Improvement of Page Ranking Algorithm Based on Timestamp and Link”, International Symposiums on Information Processing, Page(s):36-40. C. Yen, Jih Hsu (2009), “PageRank Algorithm Improvement by Page Relevance Measurement”, In FUZZ-IEEE, Page(s):502-506. D. Gleich, L. Zhukov, P. Berkhin (2004), “Fast parallel PageRank: A linear system approach Technical report” Yahoo! Research Labs. Xinyue Liu and Hongfei Lin1 and Cong Zhang(2012), “An Improved HITS Algorithm Based on Page query Similarity and Page Popularity” in Journal of Computer, Page(s):130-134. J. Wan, XueBai, “An Improvement of PageRank Algorithm Based on the Time-Activity-Curve” In IEEE conference, Page(s):549-552. D. Fetterly, M. Manasse, M. Najork and A. Ntoulas (2006), “Detecting spam Web pages through content analysis,” Proc. 15th WWW Conference, Page(s):83–92. Y Asano, Y Tezuka, T. Nishizeki(2007), “Improvements of HITS Algorithms for Spam Links” In Proceeding APWeb, Page(s):479490. Yizhou Lu, Benyu Zhang, Wensi Xi, Zheng Chen, Yi Liu, Michael R. Lyu, and Wei-Ying Ma “The PowerRank Web Link Analysis Algorithm” In 13th International Conference on World Wide Web, Page(s):254-255. A. D. Sarma, A. R. Molla, G. Pandurangan, E. Upfal(2013),“Fast Distributed PageRank Computation” In 13th WorldWide conference, Page(s):11-26. AndriMirzal, Masashi Furukawa (2010),” A Method for Accelerating the HITS Algorithm” on JACIII 14(1): Page(s):89-98. S. Shi, J. Yu, G. Yang, D. Wang (2003), “Distributed Page Ranking in Structured P2P Networks” on ICPP, Page(s):179-186.

Page 32

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net An Experimental Investigation of the Hydraulic and Durability Properties of Cement Treated Permeable Basecourses (CTPBs) Abdul A. Koroma1, Victor S. Kamara2 Department of Civil and Environmental Engineering, Michigan Technological University, Michigan, 2 (Associate Professor), Department of Civil and Environmental Engineering, Namibia University of Science and Technology, Windhoek, Namibia 1

Abstract: An experimental program was conducted to determine the hydraulic and durability properties of cement treated permeable bases using three different aggregate types, two gradations types and three targeted percent air voids content of 15, 25, and 35. It was observed that the CTPBs does have appreciable coefficient of hydraulic conductivity but that both the freeze-thaw and unconfined compressive test does not give any meaningful indication of the durability of CTPBs since the differences in UCS values between the control and conditioned samples can be attributed to differences in percent air voids content rather than due to the effect of freeze-thaw environmental procedure they were subjected to. Keywords: Hydraulic, Durability, Pavement, CTPBs, Open-graded, aggregate I. Introduction The use of permeable basecourses as the principal drainage layer within a pavementâ&#x20AC;&#x2122;s structural section has been growing in use over the years due to their perceived role of drastically reducing the time the pavement is exposed to saturated conditions. Their application has led to significant reduction in the occurrence of moisturerelated distresses in both flexible and rigid pavements. However, due to the stability problems associated with pavement sections containing unbound open-graded drainage layers, many highway agencies have resorted to the use of bound open-graded basecourses. Treatment of the unbound open-graded has been done with the use of both asphalt and cement but the asphalt treated permeable basecourses has been the most prominent. However, studies have found that after years of apparently satisfactory service, distresses have been observed in some pavements with free-draining bases even when they meet open-graded specifications. It has been observed that drainage from these layers is slowing over time and there is now increasing concern as to how long the coefficient of hydraulic conductivity of open-graded base course can be maintained as the pavement ages. Also some base course materials that do meet the required gradation specification for use as free-draining bases have only produced fair to poor drainage. This has led to the observance of premature joint deterioration, faulting and cracks in pavement sections containing asphalt treated open-graded bases. In an effort to produce non-erodible and drainable base layer, many state highway agencies have moved from the traditional dense graded base course gradation specifications to more open graded base course specifications that allow for greater drainage in the pavement sub-layers. One major reason for this transition is due to the fact that dense gradations, even though they offer stiffer bases with good constructability have serious long term stability problems as a result of prolonged saturation of the pavement structural section leading the stiffness as the pavement ages (Forsyth RA, 1994). However, one prominent drawback of these open-graded specifications is producing base layers that are difficult to construct and less stable under traffic and environmental loads. In order to overcome these drawbacks, some of these open-graded materials are now being stabilized with cement. (Hansen et al. 2009). However, studies have found that after years of apparently satisfactory service, distresses have been observed in some pavements with free-draining bases even when they meet open-graded specifications. It has been observed that drainage from these layers is slowing over time and there is now increasing concern as to how long the coefficient of hydraulic conductivity of open-graded base course can be maintained as the pavement ages (Kazmierowski et al, 94). Also some base course materials that do meet the required gradation specification for use as free-draining bases have only produced fair to poor drainage. This has led to the observance of premature joint deterioration, faulting and cracks in pavement sections containing these opengraded bases (Bennet et al. 2007). As a result of these problems concerns have been raised about the durability of cement treated permeable bases (CTPBs) especially when they are subjected to cyclic environmental conditions like freeze-thaw and wetting/drying and how this might affect their contribution to pavement performance.

Page 33

Abdul A. Koroma et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 33-39

II. Experiment Programme The experimental program was carried out in two phases. A hydraulic conductivity testing program geared towards measuring the laboratory K value of the different samples and a durability testing program for measuring the resistance of the various samples to certain environmental variables like freeze-thaw and moisture damage. Aggregate Base Materials Three aggregate base materials that were used to prepare the asphalt treated open-graded samples were two natural aggregates and one recycled aggregate. The natural aggregates were Natural Gravel and Dolomite while the recycled material was Recycled Crushed Concrete. Table 1: Aggregate material properties Material type SSD, specific Natural gravel

Absorption (%)

gravity 2.65

2.62

2.80

2.8

2.53

5.3

Limestone Recycled Portland cement concrete

Stabilizing Agents For this research program Type 1 cement was utilized for the cement treated open-graded samples. Aggregates’ Gradations Two gradation types were u s e d n a m e l y t h e AASHTO #67 and MDOT 5G. These are shown in Table 2: Table 2: Aggregate specifications – percent finer by weight Sieve Size (inches)

AASHTO'S NO.67

Michigan 5G

2 1/2 2 1 1/2 1 3/4

100 100 90-100

1/2

0-90

3/8

20-55

0-10

0-5

0-8

#16 #30 #40 #50 #200

0-3

Figure 1: Grain-size distribution of Michigan stabilized 5G aggregate

Page 34

Abdul A. Koroma et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 33-39

Figure 2: Grain-size distribution of AASHTO #67 aggregate III. Mix Design and Testing Procedure for Cement Treated Permeable Bases (CTPBs) In order to make the testing program for the laboratory hydraulic conductivity yield representative results, the mix design used in this research was made as far as possible to be practically identical to that use in actual construction projects. For this research project, only one cement content i.e. 200 lb/yd3 was used and three target air voids content of 35%, 25% and 15%. These target percent air voids content were the defining criterion of the mixed design instead of the Unconfined Compressive Strength (UCS). A constant water/cement ratio of 0.36 was used for all the mixes. Table 3 gives the mix proportions for the CTPB Table 3: Mix proportion for CTPB

Cement Content 3 (lb/yd ) 200

200

Mixture ID.

Design air void content (%) 15

Based on this design mix, 1 cubic yard of cement treated open-graded material with a design percent air void of 15% will contain 200 lbs of cement, 72 lbs of water and 3622 lbs of coarse aggregate. For a 6”*12” concrete mold and batch mix for four samples, the appropriate quantities of cement, water and coarse aggregate were determined. The absorption capacity of the coarse aggregate factions was taking into consideration in arriving at the final amount of water to be added. The purpose of doing three mix types was to find a mix design that will meet minimum acceptable stability and drainability criteria. All the preliminary testing on the aggregates was done in accordance with the ASTM standards of the respective tests. Table 3.6 gives the physical characteristics of the three open-graded aggregate materials used in this research project. Four samples for each mix type were prepared leading to a total of 72 cement treated open-graded materials. The samples were tested as follows: o All four specimens for each mix type were tested for hydraulic conductivity after 1 day curing period. The specimens were returned to the curing after the hydraulic conductivity testing. o Two samples were tested for the 7-day unconfined compressive strength (UCS) o Two samples were subjected for 10 Freeze/Thaw cycles after the 7 day curing period. o After the Freeze/Thaw procedure, the specimens were weighed to determine any weight loss and then tested for UCS to determine any strength loss as a result of the F/T process. IV. Specimen Compaction Since three design air voids content were targeted, different compaction efforts were employed in order to produce samples of the required percent air void content. As a result many trial mixes were made with different compaction efforts and the corresponding percent air void content measured. The procedure was continued until a compaction effort that produced the closest percent air void content to that of the design air void content was reached. After achieving the desired degree of compaction, the compacted

Page 35

Abdul A. Koroma et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 33-39

CTPB specimens were then tested for four key performance characteristics that are in the estimation of this researcher paramount to the characterization of cement treated drainage layers. These are the unit weight, coefficient of hydraulic conductivity, unconfined compressive strength for both control and condition specimens and the effective air void content. After making several the trial compaction runs, the following level of compactive efforts were developed: 1.

CTPB mixes with air voids content of 15%- 3 layers, 25 Marshall Hammer blows per layer and rodded 25 times per layer. 2. CTPB mixes Air voids content of 25%- 2 layers, rodded 25 times per layer, 20 Marshall Hammer blows per layer. 3. CTPB mixes with air voids content of 35%- 2 layers, no rodding, 2 Marshall hammer blows per layer After compaction, the samples were allowed to set for a day and then removed from the molds, marked, weighed, tested for the coefficient of hydraulic conductivity and then wet cure for a period of 7 days. V. Testing procedure for cement-treated Open-Graded Aggregate Materials Determination of unit weights and effective air voids content of compacted CTPB samples: There are currently no standard testing procedures for measuring the four performance characteristics of compacted open-graded concrete materials. However, a procedure developed by Crouch et al (2003) to measure similar material properties for pervious concrete was used in this research to determine the compacted unit weight and the effective air voids content of the compacted CTPB samples. Even though air voids content of 15%, 25% and 35% were used in the mix design calculations, the effective air voids of the compacted samples is the parameter of interest. The effective air voids content represents the pore spaces that are available for drainage. It is calculated from the equation: Avoid % = 100(1-Gmb/Gmm) Where Gmb = bulk specific gravity of the CTPB sample; G mm = Theoretical maximum specific gravity Experimental procedure for conducting the hydraulic conductivity test was done strictly in accordance with ASTM D5068 “Determination of Hydraulic Conductivity of Porous Stone Using a Flexwall Permeameter”. For each specimen, five hydraulic conductivity measurements were made and then averaged. Standard practice also requires that the coefficient of hydraulic conductivity be reported to a base temperature of 20 degrees Celsius, the temperature for each test was recorded and correction to the hydraulic conductivity value was made using the appropriate formula. Two correction factors were applied to the measured K value in order to arrive at the corrected measured K value. Corrections were made for the hydraulic conductivity of the ceramic porous stones and for the head loss across the specimen. The hydraulic conductivity of the porous ceramic stones was measured by setting up the flexwall permeameter using the two porous stones without the specimen in it. The head loss through the specimen was a dummy sample of similar dimension as the actual specimen made of cast steel was fitted in the laboratory set up using identical procedures was used for the actual specimen. For this case it was assumed that the head loss across the dummy sample is negligible so that head loss measured will be attributed only to that of the system (Smith 2004). VI. Durability Testing For CTPBs The freeze-thaw durability testing was performed in accordance with the procedures outlined in ASTM D560. The criterion used to measure the resistance to freeze-thaw is the total weight loss after completion of the prescribed freeze-thaw cycles. In addition to the recommended weight loss criterion, the reduction /increase of the Unconfined Compressive Strength (UCS) due to the freeze-thaw conditioning was also determined. Even though some authors don’t considered the UCS as an acceptance for cement-treated permeable bases in the same way as it applies to normal concrete, for this research study though, the percent reduction in the UCS due to the environmental conditioning of freeze-thaw was considered a measure of the stability and resistance of the cementtreated samples to disintegration. . Four CTPB samples for each mix type were divided into two groups labeled “control‟ group and “conditioned‟ group. The two control specimens were then tested for UCS after coefficient of hydraulic conductivity tests have been performed. Two other CTPB specimens were conditioned and underwent freeze-thaw in accordance with the provisions of ASTM D560. After undergoing the prescribed number of F-T cycles, the samples were weighed to determine the percent weight loss. The UCS test was afterwards conducted on the ‘’conditioned’ samples. The coefficient of hydraulic test using the flexwall permeameter was however conducted on all four CTPB samples. Two notable modifications were made to ASTM D 560 with regards to the method of compaction and the number of freeze/thaw cycles. The thickness of the CTPB normally varies between 3-6 inches, with 4 inches being the most common. For this research however, a thickness of 12 inches was selected as this height was deem adequate to achieve the three targeted range of air void content. Therefore the prescribed standard proctor compaction method was not employed since it would have lead to excessive particle breakage that may have lead to erroneous tests results which are not representative of the open-graded nature

Page 36

Abdul A. Koroma et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 33-39

of these specimens. Furthermore, in accordance with ASTM specification for the freeze/thaw test 12 freeze/thaw cycles was specified. However, 10 F/T cycles were used for this testing program because according to Dempsey (1972) the number of F/T cycles chosen for the durability test should be related to geographical location, climatic conditions and position of the stabilized layer in the pavement. Since the drainage layer for most part is located just underneath the pavement surface, it is not subjected to the same freeze/thaw cycles as the subgrade and as such a less severe F/T cycles may be appropriate for the drainage layer. The number of F/T cycles was also reduced to 10 in order to reduce the t ime for the experimental program with the assumption being that 10 F/T cycles can give a meaningful value of the resistance of these materials to freeze-thaw. Table 4: Coefficient of hydraulic conductivity of CTPB Samples Nominal Cement W/C content (lb/yd3)

Specimen

K (cm/s)

RC_5G_15

200

0.36

1.02

RC_5G_25

200

0.36

1.50

RC_5G_35

200

0.36

2.66

RC_67_15

200

0.36

0.80

RC_67_25

200

0.36

1.14

RC_67_35

200

0.36

2.17

NG_5G_15

200

0.36

1.22

NG_5G_25

200

0.36

2.06

NG_5G_35

200

0.36

3.35

NG_67_15

200

0.36

1.06

NG_67_25

200

0.36

1.39

NG_67_35

200

0.36

2.88

DL_5G_15

200

0.36

1.29

DL_5G_25

200

0.36

2.53

DL_5G_35

200

0.36

3.84

DL_67_15

200

0.36

1.01

DL 67_25

200

DL_67_35

200

0.36 0.36

2.04 2.96

Table 5: Weight and Strength Losses per sample Sample ID

RC_5G_15_1 RC_5G_15_2 RC_5G_15_3 RC_5G_15_4 RC_5G_25_1 RC_5G_25_2 RC_5G_25_3 RC_5G_25_4 RC_5G_35_1 RC_5G_35_2 RC_5G_35_3 RC_5G_35_4 RC_67_15_1 RC_67_15_2 RC_67_15_3 RC_67_15_4 RC_67_25_1 RC_67_25_2 RC_67_25_3 RC_67_25_4 RC_67_35_1 RC_67_35_2 RC_67_35_3 RC_67_35_4

Weight before F/T (g) 16700 16712 16705 16710 16214 16220 16226 16210 15980 15983 15977 15986 16728 16725 16730 16725 16245 16248 16250 16243 15992 15984 15987 15990

Measured % air voids content

Weight after F/T (g) 16704.6 16709.8 16225 16210 15976 15986 16729 16723 16249 16241 15985 15988

Weight loss (%) 0 0 0.006 0.0 0.006 0.0 0.006 0.012 0.006 0.012 0.012 0.012

UCS of control samples (7-day strength) psi 827.3 819.6

UCS of condition samples, psi

Strength loss (%)

832.4 824.5

0.6 0.6

735.2 740.4

0.8 0.4

510.7 512.3

1.3 0.4

915.4 918.4

0.6 0.3

835.3 836.4

1.0 0.5

708.7 726.3

1.0 0.4

729.1 737.6

504.2 510.3

909.6 915.3

827.2 832.5

701.8 723.4

Page 37

Abdul A. Koroma et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 33-39

Table 6: Weight and Strength Losses per sample (Cont.) Sample ID

NG_5G_15_1 NG_5G_15_2 NG_5G_15_3 NG_5G_15_4 NG_5G_25_1 NG_5G_25_2 NG_5G_25_3 NG_5G_25_4 NG_5G_35_1 NG_5G_35_2 NG_5G_35_3 NG_5G_35_4 NG_67_15_1 NG_67_15_2 NG_67_15_3 NG_67_15_4 NG_67_25_1 NG_67_25_2 NG_67_25_3 NG_67_25_4 NG_67_35_1 NG_67_35_2 NG_67_35_3 NG_67_35_4 Sample ID

DL_5G_15_1 DL_5G_15_2 DL_5G_15_3 DL_5G_15_4 DL_5G_25_1 DL_5G_25_2 DL_5G_25_3 DL_5G_25_4 DL_5G_35_1 DL_5G_35_2 DL_5G_35_3 DL_5G_35_4 DL_67_15_1 DL_67_15_2 DL_67_15_3 DL_67_15_4 DL_67_25_1 DL_67_25_2 DL_67_25_3 DL_67_25_4 DL_67_35_1 DL_67_35_2 DL_67_35_3 DL_67_35_4

Weight before F/T (g) 16825 16800 16810 16814 16735 16724 16700 16716 16230 16253 16310 16285 16835 16840 16827 16830 16742 16750 16734 16753 16315 16325 16314 16333 Weight before F/T (g) 16843 16840 16852 16842 16831 16835 16825 16820 16802 16796 16810 16805 16852 16850 16864 16870 16837 16830 16831 16828 16814 16800 16803 16811

Measured % air voids content

Weight after F/T (g) 16810 16813 16698 16715 16310 16285 16825 16830 16734 16752 16314 16333 Weight after F/T (g) 16851 16842 16825 16820 16809 16803 16864 16869 16831 16827 16802 16810

Weight loss (%)

UCS of control samples, psi

0.0 0.006 0.012 0.006 0.0 0.0 0.012 0.0 0.0 0.006 0.0 0.0

621.8 604.5

Weight loss (%) 0.006 0.0 0.0 0.0 0.006 0.012 0.0 0.006 0.0 0.006 0.006 0.006

UCS of condition samples, psi

Strength loss (%)

623.2 606.6

0.2 0.3

485.9 494.1

0.4 0.3

332.6 320.4

0.9 0.5

677.9 663.8

0.4 1.0

571.1 582.4

0.4 0.3

433.4 423.8

0.5 0.6

483.8 492.4

329.6 318.7

675.3 657.3

569.1 580.4

431.3 421.4

UCS of control samples, psi (7-day strength) 880 868.4

UCS of condition samples

Strength loss/gain (%)

883.5 876.3

0.4 0.9

746.7 756.6

0.6 0.4

577.3 576.8

0.3 0.5

968.1 971.3

0.1 1.1

816.3 811.4

0.4 0.2

716.2 717.4

0.6 0.3

742.4 753.6

575.5 574.2

969.4 960.4

813.2 809.4

712.0 715.3

VII. Conclusion From the results of laboratory testing to determine the hydraulic, mechanical and durability characteristics of cement treated open-graded aggregate base materials, the following trends were observed: 1. All three mix types with target % air voids of 15, 25 and 35 produced coefficient of hydraulic conductivity that satisfied AASHTO minimum K value of 1000 ft/day for good drainage. This means that all three design mixes can be used to provide a drainage layer that can drain within 2 hours 50% of infiltrating moisture within the pavement structural system. 2. From a material perspective, limestone mixes have the highest K values followed by those of recycled concrete aggregate and natural gravel being the least. However the differences in K value between the K values of limestone mixes and those of natural gravel is less that an order of magnitude. Since all mix types meet the minimum accepted K value, these differences in K values between the materials is not significant enough as to warrant the selection of one material over the other on the basis of hydraulic conductivity alone.

Page 38

Abdul A. Koroma et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 33-39

3. From a gradation perspective, MDOT 5G gradation samples containing the larger size fractions produced higher K values for all the mix types than mixes made from AASHTO # 67 gradation. Again the difference in K value here is less than an order of magnitude even though the MDOT gradation is more open-graded gradation with a D10 value almost 1.5 times that of AASHTO #67 gradation. One possible explanation for this is that MDOT 5G samples because of their larger size fractions require small amount of aggregates than the AASHTO #67 for the same volume. As a result the MDOT 5G samples will contain larger pore spaces between the aggregate particles which leads to a higher void ratio and subsequently higher measured coefficient of hydraulic conductivity than the AASHTO #67 samples. 4. Unconfined Compressive strength (UCS) results as expected are a direct function of the % air void content. For all mix types, the lower % air void content samples provided the highest UCS values. One noticeable trend though is that the difference in UCS values between a 15% and 35% air void content sample of the same material and gradation is about 25%. This is quite a significant amount especially when considering the stability of the drainage layer under the combined action of traffic and environmental loads. 5. Results from the durability testing showed that percentage weight loss is less than the established criterion of 5% for a stabilized layer underneath a concrete pavement. The maximum percent weight loss was 1%, which is far less than the prescribed 5%. This can be an indication that freeze-thaw durability criterion may not be an appropriate durability performance test for cement-treated open-graded bases. 6. Furthermore, the difference in UCS values between the control and conditioned samples can be attributed to strength gain due to age rather than the freeze-thaw conditioning process.

Measuring the K of CTPB using flexwall permeameter

VIII. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

CTPB samples after UCS testing

References

American Association of State Highways Transportation Officers ( AASHTO) T85 2010. Standard Method of Test for Specific Gravity and Absorption of Coarse Aggregate American Association of State Highways Transportation Officers (AASHTO) T99 2010.Standard Method of Test for Moisture-Density Relationship. American Society for Testing and Materials (ASTM )D560 .Standard Test Methods for Freezing and Thawing of Compacted SoilCement Mixtures. Annual Book of ASTM Standards, Vol.08.12.09. American Society for Testing and Materials (ASTM) D5084. Standard Test Methods for Measurement of Hydraulic Conductivity of Saturated Porous Materials Using a Flexwall Permeameter. Annual Book of ASTM Standards, Vol.08.12 2007. Crouch L, Cates M, Dotson V, Honeycutt K, Badoe D. “Measuring the Effective Air Void Content of Portland Cement Pervious Pavements” Journal of Cement, Concrete, and Aggregates. 25: 23-30, 2003. Dempsey B. J. “Durability Testing of Stabilized Bases” Interim Report., 1972 Forsyth RA .. “Asphalt Treated Permeable Base: Its evolution and Application” National Asphalt Pavement Association, Report Number Q117:23-30, Lanham, MD, 1994. Kazmierowski T. J, Bradbury A, Hajek J.. “ Field Evaluation of Various Types of Open-Graded Drainage Layers” Journal of Transportation Research Record 1434: 29-36, 1994. Marks A.. “Pervious Concrete Pavement- How Important is Compressive Strength” Journal of Green Building, 3:36-43, 2008. Mayrberger T, Hodek RJ. “Resilient Modulus at the Limits of Gradation and Varying Degrees of Saturation” Final Report Research Report RC-1497. Michigan Department of Transportation, Lansing, MI. 2007 Gupta J, Heydinger A, Randolph B.. “Permeability & Stability of Base and Subbase Materials” Ohio Department of Transportation (ODOT) Research report, 1999. Smith N. “Permeability of Pervious Portland Cement Pavement” Master of Science Thesis, Tennessee Technological University, 2004. R. Tangpithakkul, “Study of Permeability of Base Materials. Master’s Thesis, Ohio University, OH. 1997

IX. Acknowledgments The main author acknowledges the immense contribution of Dr. Ralph J. Hodek (Associate Professor) of the Department of Civil and Environmental Engineering, Michigan Technological University, USA as the PhD supervisor.orr

Page 39

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Effects of Ni and Cd contaminated fish meat on chromosomal aberrations and sperm morphology of Swiss albino mice Mustafa S.Al-Attar*and Rezan O. Rasheed** *Department of Environment, College of Science, University of Salahaddin, Erbil, Kurdistan, IRAQ **Department of Biology, School of Science, University of Sulaimani, Sulaimani, Kurdistan, IRAQ _____________________________________________________________________________________ Abstract: two heavy metals namely Ni and Cd were determined in a freshwater fish, Barbusxanthopterus using 700-atomic absorption: the highest values werefound to be (82.112 µg g-1), and (4.702 µg g-1), respectively. Flesh of this fish was fed to micein two groups for each two periods (1 and 2 months)to monitor chromosomal aberrations and sperm morphology. In Group (1) first period: the ring chromosome (3.2 ± 0.200) and centromeric gap (2.9 ± 0.314) have the highest value of aberration type, while the lowest value was chromatid break (0.0 ± 0.000), in second period: the centromere gap (4 ± 0.333) and chromatid gap (3.4 ± 0.266) have the highest value of aberration type, while the lowest value was polyploidy (0.6 ± 0.266). In Group (2) first period:the highest value of abnormality type was sperm without tail (4.9 ± 0.276) swollen head sperm (4.1 ± 0.622) and sperm without hook (3.9 ±0.433) , while the lowest value was blunt hook sperm (0.0 ±0.000).In second period: the highest value of abnormality type was defective head sperm (4.2± 0.416) and sperm without tail (3.7 ± 0.488), swollen head sperm (3.7 ±0.667) ,and the lowest value was blunt hook sperm (1.0 ±0.421). Key words: fishes, Ni, Cd, mice ______________________________________________________________________________________ I. Introduction Human activities, such as industrial and traffic emissions and various land-use practices may increase heavy metal loading into aquatic ecosystems [1]. Tulonenet al.,(2006) reported that in aquatic systems, the natural concentrations of metal ions are principally dependent on the ambient distribution, weathering and leaching of these elements from the soil in the catchments area. They also stated that heavy metals are carried to the lakes through atmospheric deposition and/or discharge. The characteristics of water, such as acidity or the amount of organic matter, are known to be important factors in determining the fate of heavy metals in lakes. It is well known that the heavy metals from man-made pollution sources are continually released into aquatic ecosystem. The contamination of heavy metals is a serious threat because of their toxicity, long persistence, bioaccumulation and biomagnifications in the food chain [2] (Authman, 2008). [2] Authman (2008) also believe that fishes can be considered as one of the most significant biomonitors in freshwater systems for the estimation of metal pollution concentration. In addition, fishes are located at the end of the aquatic food chain and may accumulate metals and pass them to human beings through food causing chronic or acute diseases (AlYousifet al.,2000 )[3]. The heavy metal concentration in fish tissues reflects past exposure via water and/or food and it can demonstrate the current situation of the animals [4] (Birungiet al., 2007). It was revealed from the studies from both field and laboratory experiments that accumulation of heavy metals in a tissue is mainly dependent upon water concentrations of metals and exposure period; although some other environmental factors such as salinity, pH, hardness, and temperature play significant roles in metal accumulation. Ecological needs, sex, age, size, feeding habits as well as biological conditions of the fish affect the heavy metals accumulation in their tissues[5] (Canli and Atli2003). Due to the deleterious effects of metals on aquatic ecosystems, it is necessary to monitor their bioaccumulation, because this will give an indication of the temporal and spatial extent of the process, as well as an assessment of the potential impact on organism health [6](Fernandes et al., 2007 ) and in order to check for those hazardous to human health[2] (Authman, 2008).Recently, [7]Sofi(2013) found administration of cadmium chloride generally produced chromosomal aberrations as well as sperm abnormalities compared to the control. The aim of this study was to determine the levels of toxic metals (Ni and Cd) in the water, and fishes and to determine their potential effects on chromosomal aberration, sperm abnormalities and on mice fedflesh of Barbusxanthoperus(Gatan) contaminated with (Ni and Cd). II. Material and methods Local Fish Samples were collected from some shops inErbil and Sulaimani cities, and then analyzed for determination of Ni and Cd, by 700-atomic absorption spectrophotometer [8] (APHA,2005). Adult male laboratory mice Musmusculusstrain BALB/c (8-10 weeks) in age, used in the present study. Diet containing prepared fish muscle instead of protein was produced, and then was given to two groups of mice for each two periods (1 and 2 months). The control groups (10 for each group) were taken the standard diet for the same periods. Chromosomal preparations from bone marrow cells were done by standard method of [9]

Page 40

Mustafa S.Al-Attar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 40-44

(Evans et al., 1964). All slides were evaluated blindly at 1000X magnification for chromosomal aberrations. Hundred well spread complete metaphases were scored per each slide. Sperm was taken from epididymis using method of [10] (Karanowska, 1976) and [11] (Wyrubek and Bruce, 1978). The prepared slides underwent microscopic examination to determine number of normal sperms and abnormal sperm out of 100. III. Statistical analysis Statistical analysis was carried out using statistically available software (SPSS version 17). Comparisons between groups were made using one-way analysis of variance (ANOVA) in combination with Duncan t-test post hoc analysis. Duncan t-test treats one group as a control, and compares all other groups against it. P values <0.01 were considered significant. IV. Results and discussion As indicated in the Table (1), the highest value of Ni of Gatan, was (82.112 µg g-1), while the lowest value was (29.557 µg g-1). The highest value of Cdof the same fish was (4.702 µg g-1), but the lowest value was (1.187 µg g-1). The heavy metals in this fish may depend on their foodintake especially food chain of fish which is not known yet. Other possibility is their capability of ions absorption from surrounding water which again needsin situ investigation. The third possibility is the intrinsic physiology of fish in itspond, especially metabolic activity. Anyhow, pollution by heavy metals is reported from other regions of Iraq, as bioaccumulation of copper, cadmium, lead and zinc is reported in Barbusbelayewi and Barbusgrypus in Diyalariver at Rustemyia an area where municipal and industrial sewage is thrown into the river [12](Coad, 2010).[1]Tulonenet al.,(2006) found that metal concentration in perch(Percafluviatilis) were higher in a humic and acid lake than in a slightly humic lake partly be explained by varying dietary regime of perch. Anyhow, further investigation is required to explore the diet of the present fish, B. xanthopterus especially at the site of catchment. Table (1): showing the amount of heavy metals (Ni and Cd) represented in µg present in the muscles of Gatan, Barbusxanthopterus examined. Heavy metal

Fish types

Minimum ( µg gm-1 )

Maximum ( µg gm-1 )

Barbusxanthopterus,

29.557

82.112

Barbusxanthopterus,

1.187

4.702

A. Effects of fish meat on chromosomal aberration in bone marrow cells of male albino mice The results of chromosomal aberrations were summarized in Table (2) which indicate the significant differences (P<0.01) between treated groups with fish meat and negative control (PBS) in total normal chromosome, total abnormal chromosome, centromeric gap, centromeric break, chromatid gap, chromatid break, ring chromosome, dicentric chromosome, pulverization and polyploidy (Figure 1), but there were no significant difference (P<0.01) in chromatid break in (G1) (in first period) and pulverization in (G2) (in second period) when compared with the negative control.The ring chromosome (3.2 ± 0.200) and centromeric gap (2.9 ± 0.314) have the highest value of aberration type in (G1), while the lowest value was chromatid break (0.0 ± 0.000). While the centromere gap (4 ± 0.333) and chromatid gap (3.4 ± 0.266) has the highest value of aberration type in (G2), while the lowest value was polyploidy (0.6 ± 0.266).

Normal metaphase

Acentric fragment

Pulverization

Chromatid gap & break

Ring chromosome

Polyploidy

Page 41

Mustafa S.Al-Attar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 40-44

Figure (1): Types of structural and numerical chromosomal aberrations (1000 X)

Normal Sperm & Swollen sperm

Sperm without head & Double head sperm

Finger head sperm

Blunt hook sperm Defective head sperm Figure (2): Types of sperm abnormalities (1000 X) Table (2): Effects of fish meat on chromosomal aberrations in male albino mice (Mean ± SE) (P<0.01). Total normal Chromo.

Total abnormal Chromo.

Centromeric Gap

Centromeric Break

Chromatid gap

Chromatid Break

Ring Chromo.

Dicentric Chromo.

Pulverization

Acentric Polyploidy Fragment

Control

94.666 ± 0.333b

5.333 ± 0.333a

1 ± 0.258a

0.333 ± 0.210a

1.333 ± 0.210a

0.333 ± 0.210a

0 ± 0.000a

0.833 ± 0.166a

1.166 ± 0.166a

0.166 ± 0.166a

82.2 ± 1.209a

17.8 ± 1.209b

2.9 ± 0.314b

2.1 ± 0.233b

2.4 ± 0.371ab

0 ± 0.000a

3.2 ± 0.200b

1.4 ± 0.400ab

2.5 ± 0.307b

2.2 ± 0.290a

1.1 ± 0.214a

78.9 ± 0.993a

21.1 ± 0.993b

4 ± 0.333b

2.8 ± 0.200b

3.4 ± 0.266b

2.4 ± 0.163b

2.9 ± 0.276b

1.7 ± 0.366b

1.3 ± 0.300ab

2.1 ± 0.314a

0.6 ± 0.266a

Period Period I (2 II (1 Months Month) ) Group 2

Treatments

Note: Similar letters in each column refer to non significant difference while different letters refer to significant difference between them. Total normal sperm

Total abnormal sperm

Sperm without head

Sperm without tail

Sperm without hook

Double head sperm

Double tail sperm

Swollen head sperm

Defective head sperm

Blunt hook sperm

Control

92 ± 0.447b

8 ± 0.447a

2.333 ± 0.333a

2.5 ± 0.223a

0.667 ± 0.210a

0.166 ± 0.166a

0.833 ± 0.307a

0.666 ± 0.333a

0.333 ± 0.210a

78.9 ± 1.852a

21.5 ± 1.641b

3.1 ± 0.276a

4.9 ± 0.276b

3.9 ± 0.433b

1.5 ± 0.521ab

1.2 ± 0.326a

4.1 ± 0.622b

2.8 ± 0.249b

0 ± 0.000a

77.5 ± 1.984a

22.5 ± 1.984b

3.2 ± 0.249a

3.7 ± 0.448ab

3 ± 0.149b

2.1 ± 0.406b

1.6 ± 0.498a

3.7 ± 0.667b

4.2 ± 0.416b

1 ± 0.421a

Period I Period II (2 (1 Months) Month) Group 2

Treatments

Table (3): Effects of fish meat on sperm abnormalities in male albino mice. (Mean ± SE) (P<0.01). Note: Similar letters in each column refer to non significant difference while different letters refer to significant difference between them.

Page 42

Mustafa S.Al-Attar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 40-44

It was clear from the Table (2) that both treatment periods of fish meat were increased the quantity of chromosome aberration, but the most effective time was second period (after 2 months). It was found from experiments on male Wistar rats received NiCl 2 in the dose of 50, 150, and 300 mg/kg that chromosome aberrations were increased significantly by NiCl2 300 mg/kg as compared to the control group (p < 0.001) [13](Tarasub and tarasub,1992). Furthermore, Ni compounds have been shown to produce singlestrand breaks in cellular DNA, as well as chromosomal aberrations and DNA-protein crosslinks ([14 & 15]Patierno and Costa, 1985;Misra et al., 1993). It has been suggested that Ni was not directly involved. Furthermore, in an interesting experiments of[16] Fahmy and Aly(2000) by using cadmium chloride they found a significant increase in the percentage of chromosomal aberrations in mouse bone marrow at the dose 5.7 and 9.5 m Kg-1. B. Effects of fish meat on sperm morphology in male albino mice The results of sperm morphology were presented in Table (3) which indicated significant difference (P<0.01) between treated groups with fish meat and negative control in most parameters analyzed (total normal sperm, total abnormal sperm, sperm without tail, sperm without hook, double head sperm, double tail sperm, swollen head sperm and defective head sperm (Figure 2). The highest value of abnormality type in G(1) was sperm without tail (4.9 ± 0.276) swollen head sperm (4.1 ± 0.622) and sperm without hook (3.9 ±0.433) , while the lowest value was blunt hook sperm (0.0 ±0.000). The highest value of abnormality type in G(2)was defective head sperm (4.2± 0.416) and sperm without tail (3.7 ± 0.488) and the lowest value was blunt hook sperm (1.0 ±0.421). It was concluded from the experiments of [17]Wadi and Ahmad(1998) that lead targets testicular spermatogenesis and sperm within the epididymis to produce reproductive toxicity rather than acting at other sites within the hypothalamic-pituitary-testicular axis after administering the mice with two concentrations of lead (0.25% and 0.5%) via drinking water for 6 weeks. On the other hand Chromium dichromate was significantly decreased the body weight gain, food and fluid intake, and epididymal sperm number, but had no effect on testis weight compared to control group after treated male mice (CD-1) with 150 ppm potassium dichromate for 12 weeks as indicated by the experiments of [18] Afonneet al., (2002).In previous experiments, [17]]Wadi and Ahmad(1999) found after administering two concentrations of lead( 0.25% and 0.5%) via drinking water in mice, the low dose significantly reduced the number of sperms within epididymis while the high dose reduced the count percentage of motile sperms and increased the percentage of abnormal sperms within the epididymis. Furthermore, reduction of spermatocytes was observed after administration of cadmium chloride(0.9, 1.9, 5.7 mg Kg -1(Fahmy and Aly, 2000)[16]. The possible mechanism in heavy metal toxicity includes the DNA-protein crosslinks by the formation of oxygen radicals (Kasprzak, 1991; Lin et al., 1992; Tarasub andTarasub,1992)[,13,19, 20 & 21].

1. 2. 3. 4. 5. 6.

V. Conclusions and Recommendations: There is a positive relation between contaminated fish meat and chromosomal aberrations as well as sperm abnormalitiesin albino micecompared to the control. People must be aware for protecting environment in general and water resources in special. Monitoring the water resources periodically so as to control any change that occur in its quality. Treatment of waste water before its discharge to the lakes and streams. Monitoring Fish quality and quantity. Controlling all fish aquariums in the area. VI. References

[1] [2] [3] [4]

[5] [6] [7] [8] [9]

Tulonen, T.; Pihlsrom, M.,Arvola,L., and Rask,M.(2006). Concentration of heavy metals in Food web components of small, boreal lakes. Boreal Environment Research 11:185-194. Authman, M.M.N.(2008).Oreochromisniloticus as a Biomonitor of Heavy Metal Pollution with Emphasis on Potential Risk and Relation to Some Biological Aspects. Global Vet.2(3): 104-109. Al-Yousif, M.H., El-Shahwani, M.S. and Mal-Ghais,S. (2000). Trace Metals in liver, skin and muscle of Lethrinuslentjan fish species in relation to body length and sex. The science of the Total Environment, 256; 87-94. Birungi, Z.,Masola,B. M. ,Zaranyika,F. , Naigaga,I., and Marshall,B. (2007).Active biomonotoring of trace heavy metals using fish (Oreochromisniloticus) as bioindicator species. The case of Nakivubo wetland along lake Victoria. Physics and Chemistry of the Earth, 32: 1350-1358. Canli, M. and Atli, G.(2003). The relationships between heavy metal (Cd,Cr, Cu, Fe, Pb, Zn) levels and the size of six Mediterranean Fish species. Environmental Pollution 121(1): 129-136. Fernandes, C., Fontainhas-Fernandes,A.,Peixoto,F. and Salgado,M.A.(2007). Bioaccumulation of heavy metals in Liza saliens from the Esmoriz-Paranos coastal lagon, Portugal.Ecotoxicology and Environmental Safety, 66:426-431. Sofi, J.S.(2013). Protective Role of Selenium and /OR melatonine Against Genotoxicity of Cadmium Chloride in male albino mice, Musmusculus BALB/C. M.Sc. thesis, University of Sulaimani. APHA-American Public Health Association.(1998). Standard methods for the examination of water and wastewater: 20 thEdn, Greenberg, A.E. Clesceri, L.S. and Eaton, AD(Eds). APHAWEF and AWWa Washington DC, USA, pp.: 1193. Evans, E.P.,Breckon,G. and Ford,C.E.(1964).An air drying method for meiotic preparation from mammalian testes.Cytogenetics, 3, 289-294.

Page 43

Mustafa S.Al-Attar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 40-44 [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

Karanowska, L. (1976).” Chromosomal mutation” cited by R. Sahai, Advances in cytogenetics and their application in livestock improvement and production, N.D.R.I. Karnal, India, (35): 136. Wyrobek, A.J. and Bruce, W.R. (1978). The induction of sperm-shape abnormalities in mice and humans, in: A. Hollaender and F.J. de Serres (Eds.), chemical mutagens; principles and methods for their detection, Vol. 5, Plenum, New York, p: 257-285. Coad, B.W.(2010).Freshwater Fishes of Iraq.Pensoft.Sofia-Moscow. Tarasub, N. and Tarasub, C.(1992).Effects of Quercetin on acute Toxiocity of rat spleen and chromosome aberrations in Bone Marrow induced by Nickel chloride. Patierno, S. R., Costa, M. (1985).DNA-protein crosslinks induced by nickel compounds in intact cultured mammalian cells.Chem-Biol. Interact. 55: 75-91. Misra M., Olinski R., Dizdaroglu M., Kasprzak K.S. (1993). Enhancement by L-Histidine of Nickel(II) induced DNA-protein cross-linking and oxidative DNA base damage in the rat kidney. Chem Res Toxicol 6: 33–37. Fahmy, M.A. and Aly,F.A.(2000).In vivo and in Vitro Studies on the Genotoxicity of Cadmium Chlorider in Mice.J.App. Toxicol. 20:231-238. Wadi,S.A. and Ahmad,G.(1999).Effects of Lead on the Male Reproductive System in Mice. J. Toxi and EnvirHealth.Part A, 56(7): 513-521. Afonne, O.,Orisakwe, O.,Ekanem. I-O.,Akumka,D.D.(2002). Zinc Protects chromium-induced Testicular Injury in Mice.IndianJ.Pharmac. 34:26-31. Kasprzak, K.S.(1991).The role of Oxidative damage in metal carcinogenicity. Chem. Res. Toxicol. 4: 604-15. Lin X., Zhuang Z.X., Costa M. (1992). Analysis of residual amino acid DNA crosslinks induced in intact cells by nickel and chromium compounds. carcinog;13:1763-8. Anon, (2006).The World Bank, Dokan and Derbendikhan Emergency Hydropower.Consultancy Services for Dokan and derbendikhan Dam Inspections, Inspection report (Final), E1537.SMEC.Section A – Complementary Grouting in Dam, Derbendikhan Dam main remedial works (Phase 1), analysis of Dambehaviour, review of previous reports recommendations, Coyne et Bellier, May 1975.

Page 44

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Evaluation of the Uncertainty in Spectral Peak Location Case Study: Symmetrical Lines J. Dubrovkin Computer Department, Western Galilee College 2421 Acre, Israel Abstract: The dependences of the relative peak shifts of Gaussian and Lorentzian symmetrical doublets on the peak separation were evaluated theoretically. It has been shown that these dependences are well described by the eighth-order polynomial over the inverse of the line separation. The shifts for asymmetrical Gaussian, Lorentzian, and Voigt doublets were calculated numerically. The qualitative patterns were obtained and some abnormal phenomena were revealed. Keywords: spectroscopy; peak identification; peak location uncertainty; peak profiles. I. Introduction Peak location of spectral lines and bands is one of the most important quantitative parameters, widely employed in theoretical and applied spectroscopy [1]. For example, classical identification of unknown elements and chemical compounds is usually performed by comparing the experimentally measured peak positions with those found in the standard library. The shifts of the peak positions are indicative of intra- and intermolecular interactions and changes of the external parameters (e.g., temperature). In this connection, the evaluation of the peak location uncertainty is of exceptional importance. However, one should regard the existence of two types of errors that arise in solving this problem. One type of errors is connected with the computational procedure [2] and instrumental factors [3], while the errors of the other type are due to overlapping of the component spectra, which may cause apparent shifts [4]. Errors of the first type can be easily eliminated by simple mathematical treatment [2, 3] and careful recalibration of the spectrometer. In contrast to this, the resolution of overlapping bands is a complicated task [5]. Therefore, from the practical point of view, it seems reasonable, first of all, to evaluate the upper limits of the apparent shifts and decide whether sophisticated computer methods should be further used in this particular case. The impact of spectral overlapping on the peak positions is usually evaluated qualitatively. The peak positions of resolved maxima are often assumed to be accurate values. Quantitative evaluation of the apparent shifts of peak positions in Gaussian and Lorentzian doublets, as well as in their ideal derivatives and in those obtained numerically, was performed only for a few particular cases by means of computer modeling [6, 7]. In these works, the dependence of the shift on the parameters of the overlapping lines was not established. The goal of this work was evaluating the peak position uncertainty caused by overlapping lines. Gaussian, Lorentzian, and Voigt profiles were studied. In what follows, for the sake of simplicity, term “line” is used for short of phrase “line and band”. The standard algebraic notations are used throughout the article. All calculations were performed and the plots were built using the MATLAB program. II. Theory A. Models Consider a doublet which maxima are located at the points and , respectively: where

is the doublet line; is the parameter of the line shape; is the full line width at half-maximum; βδ/2 is the relative separation of the doublet components; is the absolute separation; is the position of the line maximum; are the relative intensity and the relative width of the second doublet line, respectively. The following line shape functions were studied in this work: a. Gaussian function [1]: where . b. Lorentzian function [1]:

where c. Voigt function [1]:

Page 45

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

where The width of the Voigt profile [8] is

are the full widths of Gaussian and Lorentzian lines, respectively.

B. Relative shift The relative shift of the line position is usually measured with respect to the line width. However, the widths of overlapping lines can be evaluated very approximately [6, 9].Therefore, in this work, we chose to calculate the shifts with respect to the separation of the doublet components, which can be readily measured visually. In this case, the relative shift of the resolved doublet peak for the -component has the form: where

is the point at which the derivative of Eq. 1 is zero:

In the neighborhood of is positive on the left, while on the right of it is negative. Unfortunately, the analytical solution of Eq. 7 exists only in the particular cases of symmetrical Gaussian and Lorentzian doublets (see Appendix B):

where the values of A ranging from and for Eqs. 8 and 9, respectively. In all other cases, Eq. 7 can be solved only numerically, but, as we have found, the dependences obtained analytically and numerically are very close. Since, for a symmetrical doublet, the merged doublet lines display the same absolute shift at the resolution limit and below The value of the relative shift, does not depend on the separation of the doublet components. For separations larger than the resolution limit, the maxima of the first and the second lines are located at and , respectively. At the resolution limit of an asymmetrical doublet, , where is located near the zero point. Below the resolution limit, the merged doublet lines appear as a single line with a shoulder (Fig. 1). If the maximum of the first line is resolved, then and, consequently, the second line manifests itself as the right shoulder of the doublet (Fig. 1a). If the maximum of the second line is resolved, then and the left shoulder relates to the first line (Fig. 1b). In both cases, the maximum of the resolved line is shifted. It has been found (Appendix 3) that for the absolute relative shift does not depend on the separation of the doublet peaks. III. Results of Computer Modelling and Discussion A. Equal-width lines ( ) It can be seen from Eqs. 8 and 9 that near the resolution limit of symmetrical doublets ( ), the relative shift is inversely proportional to a very high power of the separation of the doublet components (Fig.2). 1. On the strength of symmetry, the line shift dependences on the line separation (Fig.2) are related as: 2. Although the more intensive second line (larger R values) causes larger shifts of the first line, its own shifts decrease. Thus, if (Fig.2), < 3. Slowly decaying interfering wings of the Lorentzian lines result in larger shifts than the corresponding quickly decaying Gaussian wings (Fig. 3): Figure 1 Asymmetrical Gauss doublet and its derivatives near resolution limit.

Doublet (blue), the 1st (green), and the 2nd (red) derivatives. (a) R=3, r=1/3, δ=1.2. (b) R=1/3, r=3, δ=0.4.

Page 46

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

Owing to symmetry,

The effect described by Eqs. 16 and 17 is significantly enhanced by increasing the intensity of the interfering line. 4. In the cases of both equal-width and non-equal-width doublet lines, the plot of the dependences for the Voigt profile lies between the corresponding plots for Gaussian and Lorentzian doublets (Fig.4). 5. The shifts of doublet peaks near the resolution limit are presented in the Table. The obtained results show that the relative shift may be as large as more than 30% of the peak separation. B. Non equal-width lines ( ) 1. If then the second line, which is more intensive and wider (larger R value), causes larger shifts of the first line; however, its own shifts decrease (Fig. 5). This effect is significantly enhanced with the second line broadening. If then the first line, which is more intensive and wider (smaller R value), causes larger shifts of the second line; however, its own shifts decrease. This effect is significantly enhanced the first line widening. The abnormal behaviour of the plots (horizontal regions) is explained by Fig. 6. The plots in this figure show the positions of the zero points of the doublet first derivatives (Appendix 3, Eq. A21), which correspond to the first line maxima increasing from the resolution limit ( =1.2) to the point of the dependence discontinuity ( =1.6) (Fig. 5, ). In this case, The discussed anomaly disappears for . 2. For a given separation, the broader second line causes larger relative shifts of both lines. Such effect is most pronounced near the resolution limit of the Lorentzian doublets than of the Gaussian doublets (Table). The more intensive second line increases the first line shift; however, its own shifts decrease (Figs. 7, 8). 3. In certain particular cases, increasing the separation of the doublet lines does not lead to the gradual increase of the resolution (Fig. 9). In the open separation interval (0.2<δ<0.4), only the right peak can be observed, while in the closed interval [0.4 δ 0.48], two peaks are resolved. However, for 0.48<δ<0.60, only the left peak can be found because the value of the first derivative at the peak location is slightly different from zero. At the endpoints, both the first and the second derivatives are equal to zero (Figs. 9c, e). This abnormal behaviour may greatly complicate peak identification, particularly, in the presence of noise. Figure 2 Dependences of relative shifts on line separation for doublets consisting of equal-width lines.

(a) Gaussian, (b) Lorentzian, and (c) Voigt lines; The red curves lie below the resolution limit. The values of R are indicated next to the curves. Figure 3 Comparison of the relative shift dependences on the line separation for doublets consisting of equal-width lines.

Gaussian (G) and Lorentzian (L) lines. R = 1(a), 2(b), 3(c), 0.5(d).

Page 47

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

Figure 4 Comparison of the relative shift dependences on the line separation for doublets consisting of Gaussian, Lorentzian, and Voigt lines.

Gaussian Gaussian (G), Lorentzian (L), and Voigt (V) lines; = 1; (c) R = 1, r = 2.

(a) R = 1, r = 1; (b) R = 2, r

Table 1: The shifts of the doublet peaks near the resolution limit. 1

0.5

1/3

0.5

1/3

0.86

36.5

1.12

1.7

14.3

1.22

0.56

9.8

1.12

14.3

1.7

1.22

9.8

0.56

0.60

33.5

0.94

2.6

17.4

1.12

0.94

14.1

0.94

17.4

2.6

1.12

14.1

0.94

0.84

27.5

1.12

1.7

10.4

1.20

0.72

12.1

1.12

10.4

1.7

1.20

12.1

0.72

0.64

7.8

9.3

0.56

12.2

33.3

0.92

0.01

12.8

0.70

6.5

3.4

0.74

4.5

1.9

0.56

19.7

5.4

0.40

34.2

33.3

0.70

1.8

15.9

0.72

15.6

1.4

0.82

14.7

0.62

0.64

9.4

8.3

0.54

14.4

28.2

0.88

0.33

14.1

0.72

9.4

2.8

0.76

11.7

1.5

0.48

4.6

6.1

0.44

5.9

14.1

0.40

12.2

25.0

0.52

3.1

2.7

0.54

2.7

1.7

0.48

18.3

3.1

0.38

20.3

10.4

0.32

25.2

25.0

0.60

15.5

0.95

0.68

14.2

0.46

0.48

13.2

5.8

0.44

8.5

14.0

0.40

11.1

19.6

0.56

8.0

2.2

0.60

8.5

1.3

1.26

9.6

11.2

1.40

3.4

6.5

1.46

2.0

6.3

1.10

33.3

16.4

1.82

14.6

0.01

1.10

5.6

25.9

1.42

1.4

18.2

1.62

0.64

17.1

0.80

33.3

34.2

1.40

15.9

1.8

1.24

8.9

15.2

1.44

2.8

9.1

1.52

1.5

11.3

1.08

33.6

14.1

1.76

13.6

0.33

1.40

6.3

9.9

1.52

2.8

5.7

1.58

1.7

4.9

1.28

14.4

11.1

1.20

25.0

12.2

1.42

3.2

21.5

1.76

1.0

20.8

2.00

0.48

17.6

1.12

10.6

23.9

0.94

25.0

30.9

1.44

5.7

12.0

1.64

2.3

10.0

1.76

1.3

10.3

1.28

13.9

11.8

1.06

25.1

17.7

For each the calculated parameters for Gaussian, Lorentzian, and Voigt doublets are shown in the 1st, 2nd, and 3rd rows, respectively. Small errors are highlighted in bold.

Page 48

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

Figure 5 Comparison of the relative shift dependences on the line separation for doublets consisting of non-equal-width lines.

Gaussian (a), Lorentzian (b), and Voigt (c) lines ( values of R are indicated next to the curves.

Red curves lie below the resolution limit. The

Page 49

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

Figure 6 Asymmetrical Gaussian doublets and its derivatives at and above the resolution limit.

Doublet (blue), the 1st (green), and the 2nd (red) derivatives. R=3, r=3. Figure 7 Comparison of the relative shift dependences on the line separation for doublets consisting of the non-equal-width lines.

Gaussian (a) and Lorentzian (b) lines. The values of r are indicated next to the curves.

Page 50

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

Figure 8 Comparison of the relative shift dependences on the line separation for doublets consisting of the non-equal-width lines.

Gaussian (a) and Lorentzian (b) lines. The values of r are indicated next to the curves. Figure 9 Anomalous effects of the identification of the peak positions for Voigt doublets.

Doublet (black), the 1st (blue), and the 2nd (red) derivatives.

Appendix 1. Calculation of the peak shifts a symmetrical Gaussian doublet For a symmetrical Gaussian doublet Eq. 7 has the following form after eliminating the common factor: Since

and

, it follows from Eq. A1 that

Page 51

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

where , The trivial solution of Eq. A2, is the central point of the doublet. The other roots of transcendental Eq. A2 can be found using a most precise approximation of the exponents based on continued fractions [10]: where A3 into Eq. A2, the following cubic equation over z=

By substituting Eq. is obtained:

Since the direct solution of Eq. A4 is very complicated, we used quadratic approximations in the range of the variable, being of practical relevance: where Coefficients and were calculated by reducing the polynomial Lanczos power series expansion [11]. Thus we arrive at the final equation: where The positive solution of Eq. A6 is

degree using the telescopic shift of the

where The relative shift (Eq. 6) now has the form: A simplified form of Eq. A8 was obtained using polynomial approximation (the MATLAB function polyfit):

where 2. Calculation of the peak shifts for a symmetrical Lorentzian doublet After eliminating the common factor, Eq. 7 for a symmetrical Lorentzian doublet

The trivial solution of Eq. A10, found from the equality

has the form:

is the central point of the doublet. The other roots of Eq. A10 can be

where and The positive solution of Eq. A11 is where and The relative shift (Eq. 6) is A simplified form of Eq. A13 was obtained using polynomial approximation (the MATLAB function polyfit): where 3Calculation of the zero point of the first derivative of an asymmetrical doublet in certain particular cases In the cases of asymmetrical Gaussian and Lorentzian doublets, Eqs. A1 and A10 have the forms:

respectively.

Page 52

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013-February, 2014, pp. 45-53

The above two equations have trivial solutions if two conditions are fulfilled:

The solution of Eq. A17 is Substituting Eq. A19 into Eq. A17, we obtain: From Eq. A19 it follows that

For symmetrical doublets first doublet peak. In this case, relative shift

then

i. e., the zero point may correspond to the

does not depend on the separation of the doublet peaks. For example, if and 3, then and 0.25, respectively. For , Eq. A21 only changes its sign ( Therefore, the zero point may correspond to the second doublet peak and/which leads to

References B. K. Sharma, Spectroscopy. 19th Ed. India, Meerut-Delhy: Goel Publishing House, 2007. L. Smeller, “How precise are the positions of computer-determined peaks?”, Appl. Spectr., 1998, vol. 52, pp. 1623-1626. H. Witjes, M. Pepers , W.J. Melssen and L.M.C. Buydens, “Modelling phase shifts, peak shifts and peak width variations in spectral data sets: its value in multivariate data analysis ”, Anal. Chim. Acta, 2001, vol. 432, pp. 113-124. [4] Soo Ryeon Ryu, Isao Noda and Young Mee Jung, “What is the origin of the positional fluctuation of spectral features: true frequency shift or relative intensity changes of two overlapped bands ?”, Appl. Spectr., 2010, vol. 64, pp. 1017-1021. [5] T. O’Haver, Fourier Deconvolution, http://terpconnect.umd.edu/~ toh/spectrum/Deconvolution.html. [6] J. M. Dubrovkin and V. G. Belikov, Derivative Spectroscopy. Theory, Technics, Application. Russia: Rostov University, 1988. [7] G. Talsky. Derivative Spectrophotometry. Low and Higher Order. Germany, Weinheim: VCH Verlagsgesellschaft , 1994. [8] J. J. Olivero and R.L. Longbothum, “ Empirical fits to the Voigt line width: A brief review”, J. Quant. Spectroscopy and Radiative Transfer , vol. 17, 1977, pp. 233–236. [9] V. A. Lóenz-Fonfría and E. Padrós, “Method for the estimation of the mean lorentzian bandwidth in spectra composed of an unknown number of highly overlapped bands ", Appl. Spectr., 2008, vol. 62, pp. 689-700. [10] L. I. Turchak and P. V. Plotnikov, Fundamentals of numerical methods, Russia, Moscow: Fizmatlit, 2003. [11] C. Lanczos, Applied analysis, New York: Dover Publication, Inc, 1988. [1] [2] [3]

Page 53

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A ROBUST APPROACH FOR OBJECT TRACKING BASED ON PARTICLE FILTER AND OPTIMIZED LIKELIHOOD Amr M. Nagy1, Ali Ahmed2 and Hala H. Zayed3 1,3 Faculty of Computers and Informatics Benha University Benha, Egypt 2 Faculty of Computers and Information Menofia University Shubin el Kom, Menufia, Egypt Abstract: Robust tracking of non-rigid objects is a challenging task. Particle filter is a powerful tool for vision tracking based on sequential Monte Carlo framework and proved very successful for non-linear and nonGaussian estimation problem. This paper proposes a tracking algorithm based on particle filter and optimized Likelihood. Colour distributions are applied as they are robust to partial occlusion, rotation, scale invariant and computationally efficient. As the colour of an object can vary over time dependent on the illumination, the target model is adapted during temporally stable image observation. Particle filter approximates a posterior probability density of the state by using samples which are called particles. Here, the state is treated as the position of the object and the weight is considered as the likelihood of each particle. For this likelihood, we calculate the similarity between the colour histogram of the tracked object and the region around the position of each particle by using Bhattacharya distance. To enhance the results, a new parameter is multiplied by the previous likelihood to increase the particles weight. The system proves to be robust against problems of partial occlusion, full occlusion and illumination changes. Finally the mean state of the particles is treated as the estimated position of the object. The correctness as well as validity of the algorithm is demonstrated through the experiments results. Keywords: Particle Filter; Colour Histogram; Nonlinear/NonGaussian; Object Tracking; Optimized Likelihood I.

Introduction

Tracking is an essential step in many computer vision related applications. Object tracking is the task of detecting and following an object of interest, over period of time. Vision based tracking system detects and tracks objects in a sequence of images. Object Tracking is required by many vision applications such as surveillance [1], human computer interfaces and video communications/ compression. To define your Object of interest it depends on the specific application at hand. For example, in a building surveillance application, targets may be people, whereas in an interactive gaming application may be hands or the face of a person. Numerous approaches have been proposed to improve the performances of target tracking, which have achieved significant improvement in the past decades. They can be roughly classified into two categories: deterministic methods and stochastic methods. Deterministic methods typically track the object by performing an iterative search for a similarity between the template image and the current one. The algorithms which utilize the deterministic method are background subtraction [2,3] inter-frame difference [4,5], optical flow [6], skin colour extraction [7,8] and so on. On the other hand, the stochastic methods use the state space to model the underlying dynamics of the tracking system such as Kalman filter [9], particle filter [10–14] Hybrid Blob and Particle Filter Tracking Approach for Robust Object Tracking, Object Tracking Using Hybrid Mean Shift and Particle Filter and Hybrid Iterated Kalman Particle Filter [15-17]. Probabilistic methods have become popular among many researchers. The Kalman filter is a common approach for dealing with target tracking in a probabilistic framework, but it cannot resolve a tracking problem where the model is nonlinear and non-Gaussian. The extended Kalman filter can deal with this problem, but still has a problem when the nonlinearity and non-Gaussian cannot be approximated accurately. Recently, the particle filter method, a numerical method that allows finding an approximate solution to the sequential estimation has proven very successful for nonlinear and non-Gaussian estimation problems. It approximates a posterior probability density of the state such as the object position by using samples which are called particles. An important issue in particle filtering is the selection of the proposal distribution function. In general, it is hard to design such proposals. Now many proposed distributions have been proposed in the literature. For example, the prior, the EKF Gaussian approximation and the UKF proposal are used as the proposal distribution for particle filter [16–18]. In this paper, a new proposal distribution generating scheme for the particle filtering framework is proposed. The algorithm obtained is named as particle filter with optimized likelihood. In this algorithm, we propose to use such a particle filter with color-based image features. Color histograms in particular have many advantages for tracking IJETCAS 14-111; © 2014, IJETCAS All Rights Reserved

Page 54

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

non-rigid objects as they are robust to partial occlusion, rotation, scale invariant and are calculated efficiently. A target is tracked with a particle filter by comparing its histogram with the histograms of the sample positions using the Bhattacharyya distance, which consequently improves the performance of particle filters to estimate the new state of the tracked object. II. Basic Particle Filter. The state equation and measurement equation of the dynamic system are described as follows:

Where xk denotes the system state at time k, and yk denotes the observation at time k. vk and uk are the process noise and measurement noise at time k respectively (they obeys the independent and identical distribution). The state model f(.) and observation model h(.) are known and at least one non-linear. The state equation (1) characterizes the state transition probability of the system P(xk|xk−1), and measurement equation (2) characterizes the likelihood probability P(yk|xk). From the perspective of Bayesian filter, given that the initial state x0 is P(x0|y0) ≡ P(x0), the state transition probability P(xk|xk−1) and likelihood probability P(yk|xk) the problem-solving core is to estimate the posterior probability density function (PDF) P(xk|yk). The particle filter is the Bayesian filter’s variety. It uses a set of weighted samples to approximate the posterior probability density function

The particle filter algorithm has three important steps: particle production (important sampling), weight computation and resampling. Step 1 Produce particle (important sampling)

Step 2 Compute weight and normalize weight

Step 3 State estimate

Steps 4 Resample. Duplicate the high weight particle and get rid of the low weight one from the particle set new particle set

, obtain the

. III.

Re-sampling

This step involves discarding samples that have low importance and reassigning weights to the remaining particles. Various approaches have been suggested in the literature for carrying out this step. IV.

The Proposed Tracking System

Before talking about our proposed algorithm (Particle Filter with optimized likelihood), firstly the Color Based Distribution and the Particle filter initialization are introduced.

Page 55

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

A. Color Based Distribution Assume that the distributions are discretized into m bins. The histograms are produced with the function h (xi) that assigns the colour at location xi to the corresponding bin. In our experiments, to make the algorithm less sensitive to lighting conditions, HSV colour space using (8*8*8 ) bins is used to compute the histogram. Here to compute the weight of the sample set, we do not use the entire image as a measurement, but rather we compute the colour histogram inside the ellipse that is specified by the state vector. After we compute the histogram, we use the Bhattacharyya distance to compute the similarity between the two colour histograms p = p(u), u = 1, ....,m which taken from the first frame and q = q(u), u = 1, ....,m which taken from the next frame. Bhattacharyya distance are calculated using the following equation

Where

From this equation, when the is large this indicate that the distribution is more similar. If perfect match and we have a two identical histograms

= 1 this indicate a

B. Particle filter initialization For the initialization of the particle filter, we have to find the initial starting values. There are three possibilities depending on the prior knowledge of the target object: manual initialization, automatic initialization using a known histogram as a target modal or an object detection algorithm that finds interesting targets. Whatever the choice, the object must be fully visible, so that a good colour distribution can be calculated. C. Particle filter with optimized likelihood. Our proposed algorithm named particle filter with optimized likelihood. It inherits the excellent properties of the colour histogram, which make it very attractive for the generation of proposal distribution within the particle filtering framework. Our proposed tracking system framework uses the Bhattacharyya distance to update the priori distribution calculated by the particle filter. Before applying the tracking algorithm, we detect an interested object manually to segment it from the background scene. For a new object entering at time instance k, the system initializes its system state . Commonly used appearance models are colours values of fitted ellipse (colour matrices), and compact summarization of colour distribution such as histograms. The position ( , ) is coordinate of an object in image plane. The velocities and are initialized as zeros. The sizes ( , ) are the length of the major axis and the minor axis of the ellipse fitted on the visual object and the corresponding scale change. The sample set is propagated through the application of a dynamic model

Where A defines the deterministic component of the model and is a multivariate Gaussian random variable. In our application we currently use a first order model for A describing a region moving with constant velocity and and scale change . After we propagated the particles according to the system modal. The weight is considered as the likelihood of each particle. For this likelihood, we use the Bhattacharyya distance to compute similarity of the colour distribution of the tracked object and the region around the position of each particle. After we compute the likelihood, we multiply it with the proposed parameter Îą. It consists of multiplication of number of particles, dimensional of the state vector and the number colours of histogram. We use this parameter to increase the weight of each particle. The weight

of the i-th state

is calculated as

Page 56

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

where and q are the color histogram of sample and target ,respectively. From this equation, if we obtained a small Bhattacharyya distance, this indicates that we have large weight. During the resample step of the particle filter, samples with high weight are chosen several times, leading to identical copies, while others with relatively small weights are removed. Figure 1 Flow chart of our proposed algorithm

Experimental Results.

To demonstrate the improved particle algorithm, we used two different datasets. The first one for indoor dataset taken on our lab which consists of 491 frames and every frame has 480 width and 272 height and the second one for outdoor dataset (PETS2009) S2.L1 collection with scenario walking with elements sparse crowd is investigated which consists of 287 frame and every frame has 768 width and 576 height. We choose the sequences from view 001 for our evaluation due to the wide angle view in order to reduce the possibility of capturing object partially. Dataset view 001 consist more than 8 persons with similar colour properties walking from various directions. A. Results using indoor dataset. We test the algorithm for indoor dataset with different number of particles. The tracking results for the proposed system are shown in figure (2, 3) using 100 and 300 particles respectively. The estimated trajectory shows that, the proposed algorithm could track the object under illumination changes. Figure 4, shows the estimated trajectory and the original trajectory of the object through the video sequence. When we use 300 particles in figure 4 (b), we obtain the best trajectory but in figure4 (a), with 100 particles after 300 frames, Itâ&#x20AC;&#x2122;s obvious that the estimated trajectory diverse from original trajectory.

Page 57

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

Figure 2 Tracking result for indoor dataset using 100 particles.

Figure 3 Tracking result for indoor dataset dataset using 300 particles

Figure 4 Tracking trajectories for indoor dataset using 100 and 300 particles respectively.

Page 58

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

B. Results using outdoor dataset (PETS2009 dataset). We also test the algorithm for outdoor dataset with different number of particles. The tracking results of the proposed system are shown in figure (5, 6) using 100 and 300 particles respectively. Figure 7 show the estimated trajectory and the original trajectory of the object through the video sequence. When we use 300 particles in figure 7 (b), we obtain the best trajectory but in figure 7 (a), with 100 particles after 150 frames, Itâ&#x20AC;&#x2122;s obvious that the estimated trajectory diverse from original trajectory. Also the system show best tracking result in case of partial occlusion and full occlusion as seen in figure 5 (frames 196, 197, 230, 233) and in figure 6 (196, 197, 230, 233) . Figure 5 Tracking result for outdoor dataset using 100 particles.

Figure 6 Tracking result for outdoor dataset using 300 particles.

Page 59

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

Figure 7 Tracking trajectories for outdoor dataset using 100 and 300 particles respectively.

C. System performance. We measure the performance of the proposed system using the root mean square error (RMSE) for PF with color histogram and PF with optimized likelihood. In case of indoor dataset, we obtained the RMSE 10.34 and 4.10 for 100 and 300 particles with color histogram respectively, and 9.9 and 3.26 for 100 and 300 particles with optimized likelihood respectively. Using the outdoor dataset, it gives 4.59 and 3.74 for 100 and 300 particles with color histogram respectively, and 4.38 and 2.75 for 100 and 300 particles using optimized likelihood respectively. For both experiments, the system gives high performance when using 300 particles than 100 particles. Table I Shows RMSE and time at different number of particles for indoor dataset Sequence with PF Co. histogram PF OP. likelihood

100 RMSE Time 10.34 30.32 9.90 31.17

150 RMSE Time 5.99 38.74 5.85 36.81

particles 200 RMSE Time 4.89 39.19 4.50 39.54

250 RMSE Time 4.53 40.74 4.27 41.64

300 RMSE Time 4.10 44.61 3.26 44.89

Table III Shows RMSE and time at different number of particles for outdoor dataset Sequence with PF Co. histogram PF OP. likelihood

100 RMSE 4.59 4.38

150 Time RMSE 28.26 3.98 29.01 3.95 VI.

Time 31.85 31.20

particles 200 RMSE Time 3.77 30.52 3.62 30.74

250 RMSE 3.91 2.95

Time 34.97 35.5

300 RMSE 3.74 2.75

Time 38.42 37.8

Conclusions and Future Work.

In this paper, a robust tracking algorithm is presented, which combines particle filter and optimized likelihood. The experimental results demonstrate that the proposed algorithm can effectively overcome the problems of object occlusion and can track the color target efficiently in presence of illumination changes. To obtain more accurate result, one can use multi features with particle filter such as first and second derivatives edge detection methods to enhance the tracking of the object. VII. References [1]

Kyungnam Kim,Larry S. Davis, Object Detection and Tracking for Intelligent Video Surveillance, Multimedia Analysis, Processing and Communications, pp. 265-288. 2011.

[2]

McIvor, A. M. Background subtraction techniques, Proceeding of Image and Vision Computing, 6 pages. 2000.

[3]

LIU, Y.; Haizho, A. & Xu Guangyou, Moving object detection and tracking based on background subtraction, Proceeding of Society of Photo-Optical Instrument Engineers, Vol. 4554, pp. 62-66. 2001.

[4]

Lipton, A; Fujiyoshi, H. & Patil, R.,Moving target classification and tracking from real-time video, Proceeding of IEEE Workshop Applications of Computer Vision, pp. 8-14. 1998.

Page 60

Amr M. Nagy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February 2014, pp. 54-61

[5]

Collins, R. ; Lipton, A.; Kanade, T.; Fujiyoshi, H.; Duggins, D.; Tsin, Y.; Tolliver, D.; Enomoto, N. & Hasegawa., System for video surveillance and monitoring, Technical report CMU-RI-TR-00-12, Robotics Institute, Carnegie Mellon University, 2000.

[6]

Meyer, D.; Denzler, J. & Niemann, H.,Model based extraction of articulated objects in image se-quences for gait analysis, Proceeding of IEEE Int. Conf. Image Proccessing, pp.78-81. 1998.

[7]

Cho, K. M.; Jang, J. H. & Hong, K. S., Adaptive skin-color filter, Pattern Recognition, pp. 1067-1073. 2001.

[8]

Phung, S.; Chai, D. & Bouzerdoum, A., Adaptive skin segmentation in color images, Proceeding of IEEE International Conference on Acoustics, Speech and Signal Processing, Vol. 3, pp. 353-356. 2003.

[9]

Broida, T. & Chellappa, R., Estimation of object motion parameters from noisy images, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol 8. No 1.pp. 90-99 1986.

[10] Isard M. & Blake A., CONDENSATION-Conditional Density Propagation for Visual Tracking, Intl. Journal of Computer Vision,Vol. 29,No. 1, pp. 5-28 1998. [11] Ristic, B.; Arulampalam, S. & Gordon, N., Beyond the Kalman filter: Particle filters for tracking applications, Artech House, 2004. [12] Xu Fen, Gao Ming.,Pedestrian Tracking Using Particle Filter Algorithm, International Conference on Electrical and Control Engineering., 2010. [13] Zhiqiang Wen., Zhaoyi Peng, Xiaojun Deng, Shifeng Li.,Particle Filter Object Tracking Based on Multiple Cues Fusion, Advanced in Control Engineering and Information Science.,pp. 1461-1465 2011. [14] QU Zhonga , ZHANG Qingqinga, GAO Tengfeia,Moving Object Tracking Based on Codebook and Particle Filter, International Workshop on Information and Electronics Engineering.,pp. 174-178 2012. [15] Sze Ling Tanga, Zulaikha Kadima, Kim Meng Lianga, Mei Kuan Lima,Hybrid Blob and Particle Filter Tracking Approach for Robust Object Tracking, IProceedings of the International Conference on Computational Science.,pp. 2549-2557 2010. [16] Asad, Naeem.; Tony, Pridmore., Object Tracking Using Hybrid Mean Shift and Particle Filter Algorithms: An indepth discussion on computer vision object tracking algorithms, LAP LAMBERT Academic, 2012. [17] Amr M. Nagy, Ali Ahmed, Hala H. Zayed,Hybrid Iterated Kalman Particle Filter for Object Tracking Problems, IProceedings of the International Conference on Computer Vision Theory and Applications.,pp. 375-381 2013. [18] Gordon, N.J., Salmond, D.J., Smith, A.F.M.,Novel approach to nonlinear/non-Gaussian Bayesian state estimation [J], IEEE Proceedings Radar and Signal Processing,Vol. 140, pp. 107-113 1993. [19] Arulampalam, M.S. Maskell, S., Gordon, N., Clapp, T.,A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking[J], IEEE Transactions on Signal Processing, Vol. 50, pp. 174-188 2002. [20] R Van der Merwe, A Doucet.,The Unscented Particle Filter, Advances in Neural Information Pro-cessing Systems [M], MIT, 2000. [21] PETS 2009 Bnechmark Data http://www.cvg.rdg.ac.uk/PETS2009/a.html#s2

Page 61

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Star-Mobius Cube: A New Interconnection Topology for Large Scale Parallel Processing Debasmita Pattanayak1, Devashree Tripathy2 and C.R.Tripathy3 Department of Computer Science and Engineering, VSS University of Technology, Burla, Odisha, INDIA 2 Department of Advanced Electronic Systems, Central Electronic Engineering Research Institute (CEERI), Pilani, INDIA 3 Department of Computer Science and Engineering, VSS University of Technology, Burla, Odisha, INDIA ________________________________________________________________________________________ Abstract: The interconnection topology plays a vital role in parallel computing systems. In this paper a new interconnection network topology named as Star-mobius cube (SMQ) is introduced. The various topological and performance parameters such as diameter, cost, average distance, and message traffic density are discussed. The embedding and broadcasting aspects of the new network are also presented. Based on the performance analysis, the proposed topology SMQ is proved to be a better alternative to its contemporary networks. Key words: Interconnection networks, topological parameters, broadcasting, routing, embedding. ________________________________________________________________________________________ 1

I. Introduction The rapid progress in VLSI technology has lead to the development of multiprocessor systems that constitutes large scale parallel processing. In such systems numerous processors work together to execute tasks in parallel. A parallel computing system is of two types: Loosely coupled system and tightly coupled system. In a tightly coupled system, the processors are interconnected through a shared memory. However in loosely coupled system, each processor has its own private memory. To execute a task the processors in a multiprocessor system should exchange data among themselves and the interconnection network undertakes the active role in this data exchange. The various network topologies for interconnection proposed in the literature include star[4], hypercube[8], tree, mesh and the ring[3]. Extensive research has been done on cube based networks as it has lower diameter, scalability and high embeddability. The different derivatives of hypercube have been proposed in literature. The prominent candidates among them are crossed cube[2] ,dual cube[10] , meta cube[9] , mobius cube[1] and the star cube[5] . The Star network for the long time has been in research due to its attractive properties like maximum fault tolerance and optimal broadcasting. Different performance parameters of a good network topology are small diameter, low degree, low cost, low average distance, low message density, efficient routing and broadcasting. Recently more research works are being done on product graphs. The main objective of a product graph network is to derive a new topology which comprises of positive features of both the base graphs. Many product networks such as Starcube[5], Star crossed cube[6], have been investigated The principal objective here is to design a new topology which can be better than the star graph, mobius cube, Starcube and Star crossed cube. Our proposed topology in this work is the Star Mobius cube(n, k) which is a product graph of n- star[4] and k- mobius cube[1]. The proposed network Star mobius cube inherits properties of both the star graph and the mobius cube and possesses many attractive properties as compared to its parent graph and other product graphs. Next, we compare the performance parameters of the said networks to substantiate the merits of the new network. This paper is organized as follows .The Section II describes the related work. In section III, we present the proposed topology and derive its properties. We propose algorithms for broadcasting and routing in Sections IVV respectively. In Section VI, we discuss about embedding of other networks in the proposed topology. We analyse the performance and illustrate the merits of the proposed topology in section VII. The Section VIII concludes the paper. II. Related work The hypercube [8] is considered to be the best among the various loosely coupled topologies. The most improved form of hypercube is the crossed cube [2]. The Crossed cube is better than the hypercube in terms of its diameter i.e.(n+1)/2. It is also edge pancyclic for n≥2. But the crossed cube has higher message traffic density than the hypercube. The Dual cube[10] is an another variant of hypercube. The Meta cube [9] proposed by Li, Peng and W.Chu has the least no. of links. But the embedding of metacube and dualcube in other networks is quite difficult. Paul Cull and M. Larson showed that the mobius cube[1] has effectively smaller diameter than hypercube. The n-star graph [4] is a permutation graph as it has n! nodes. The Star graph is a better alternative to the cubic networks. But the factorial increase of nodes is the main drawback of the star

Page 62

D. Pattanayak et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 62-68

graph. An interconnection Star Cube topology [5] has better diameter and average distance than the hypercube and the star network. It inherits the advantages of both the star and the hypercube. Subsequently, Star Crossed Cube[6] has been proposed with better performance measures than the Starcube. Here, the main challenge that remains to be addressed is the development of a new topology that overcomes the drawbacks of the above said topologies. III. Proposed Topology In this section, our main objective is to propose a new topology taking into account the best features of the star graph and the mobius cube. Before proceeding further, a brief discussion on the parent networks follows. 3.1 Star graph The m-dimensional Star denoted as S(m) has m! no. of nodes (x0,x1,x3.....xm-1)[4]. Each node of S(m) is represented by permutation of address bits. The address bits of each node are arranged with SWAPi operation in routing. There exists an edge between two nodes when the address bits differ by their position e.g. 123 is the first node; then SWAP2 i.e. swap between initial and second position, so the next node is 213. A Star graph has m!(m-1)/2 number of edges. But when we go for higher dimensional star graph then the number of nodes increases in a higher rate. For example, S(3) has 3!=6 no. of nodes where as S(4) has 4!=24 no. of nodes i.e.4x differ in S(3).The m-star graph is vertex and edge symmetric and has a diameter 3/2(m-1) and is a m! permutation network. Because it has a set of m-1 generators, so the degree of star graph is m-1[4].The Fig.1 illustrates a 3-dimensional Star graph. Fig. 1: Star network(dimension = 3)

Fig. 2(a): Mobius cube MQ30

Fig. 2(b): Mobius cube MQ31

3.2 Mobius cube An n-dimensional mobius cube MQn has 2n nodes [1]. An edge joins node X=x1x2x3....xn to the node Y=y1y2y3...yn if Yi satisfies one of the following condition: Yi = x1.....xi-1 i xi+1......xn if xi-1=0 (1) Yi = x1 ..... xi-1 i i+1.....xn if xi-1=1 (2) Where i is the complement of xi in (0,1). When the address bits differ only in the leftmost bit then the mobius cube known as 0-type mobius cube MQ30 and when address bits differ in all bits then it is known as 1-type mobius cube MQ31. According to the condition (1) an edge between a node and its jth node can be established if its jth value is 1. Similarly, based on second rule if jth through nth components equal to 1.The Figs. 2(a) and 2(b) show the 0-type mobius cube and 1-type mobius cube respectively. In the n-dimensional mobius cube the diameter is (n+1)/2. In mobius cube, the routing algorithm takes O(n) runtime. On comparison of topological properties the mobius cube is found better than other variants of hypercube. Mobius cube has also efficient broadcast algorithm. However 1-type mobius cube is better than 0-type mobius cube. We can construct a 4dimensional mobius cube out of figure 2(a) and 2(b) by following the above conditions 1and 2. 3.3 Proposed topology: Star mobius cube In this section, we propose the new topology Star-mobius cube. The Star mobius cube denoted as SMQ(n,k) is the product graph of m-Star S(m) and K-mobius cube MQk. Here, each node of star graph is substituted by mobius cube. The address of each node in SMQ has two parts Xi that represents star part and Yi represents MQ part {x0x1....xnyk-1yk-2......y0}. In simple logic, mobius cubes are placed on star platform. The Fig. 3 illustrates the proposed Star-Mobius cube topology SMQ(3,3) for dimension 3. 3.3.1 Topological features of Star mobius cube(n,k) This subsection derives the expressions for various topological parameters of the proposed network. a) Node (N) The total number of nodes of the network shows the network size. Theorem 1: The total number of nodes in the SMQ(n,k) graph is n! 2k. Proof: An n-star has n! number of nodes and the MQ(k) has 2k nodes. So the SMQ(n.k) which is the product graph of MQk and n- Star network shall have n! 2k nodes in all.

Page 63

D. Pattanayak et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 62-68

b) Edges or Links (E): The edges connect the nodes and communication between nodes takes place through the edges. Theorem 2: The total number of edges in the SMQ(n, k) is n! 2k-1 (k + n- 1). Proof: The SMQ(n,k) has n! MQk nodes connected to n-1 neighbours and each MQ has k2k-1 edges. Hence, to connect n! MQs in star requiers n! 2k-1 (k + n- 1) number of edges. Illustration: In SMQ(3,3) the star dimension n=3 and the mobius cube dimension k=3.There can be 6 no. of mobius cube k-1 and each mobius cube has k 2 i.e. 12 no. of links. All 6 mobius cube need to be connected in star platform. So the total no. of edges :3! *23-1(3+3-1)= 6*4*5= 120 no. of edges.

Fig. 3: Star Mobius cube c) Degree The degree is defined as the total no. of edges come out from each node. In symmetric network like Starmobius cube, each node has equal node degree. If the degree of network is high then it can connect more no. of nodes. So the requisite is higher degree network topology. Theorem 3: The degree of SMQ(n,k) is (n+k-1) . Proof: In MQk , each node’s degree is k and (n-1) edges are incident from each node of MQk . So, the degree of SMQ is k+n-1. d) Diameter The diameter of an interconnection network is defined as the maximum distance between any two nodes in the network. The distance between two nodes is the shortest path between the nodes. Obviously the diameter network topology should be low so that we can have optimal number routing steps with higher degree, otherwise the complexity may rise. Theorem 4: The diameter of the Star- mobius cube topology SMQ(n, k) is (3(n-1)/2) + (k+2)/2 k>=4 in 0-type SMQk . 3(n-1)/2 + (k+1)/2 k>=4 in 1-type SMQk . Proof: In an n-star travelling from one node to other takes (3(n-1)/2) steps. For MQk0 the diameter is (k+2)/2 k>=4 . For MQk1 the diameter is (k+1)/2 k>=4. This lower , the diameter, the better is the network. In (3(n-1)/2) + (k+2)/2 k>=4 number of hops for SMQk0 or 3(n-1)/2 + (k+1)/2 k>=4 number of hops for SMQk1 ,we can move from any node to any other node in SMQ. e) Cost The cost is a significant performance measure of any network topology. The cost deals with the communication links. The cost is the product of network degree and diameter. It should be less for a network. Theorem 5: The cost of the Star-Mobius cube network SMQ(n, k) is (n+k-1)[3(n-1)/2 + (k+2)/2] k>=4 SMQk0 . (n+k-1)[3(n-1)/2 + (k+1)/2] k>=4 SMQk1. Proof: In SMQ the cost= DegreeDiameter; we know from Theorem 3 that the degree of (SMQ) is (n+k-1) and from Theorem 4 the diameter of 0-type and 1-type mobius cube. Using both theorems we can get the expression of cost for SMQ. f) Average Distance (d) Theorem 6: The average distance (d) of the SMQ(n,k) is given by

Page 64

D. Pattanayak et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 62-68

k /3+[1-(-1/2)k]/9 +n-4+2/n+ ≤ (d ) ≤ k/3+[1-(-1/2)k]/9+1+n4+2/n+ Proof: The average distance of the n-star graph is n-4+2/n+ and for MQk0 average distance is k /3+[1-(k 1 k 1/2) ]/9 and for MQk , k /3+[1-(-1/2) ]/9 +1. Hence, the average distance of SMQ is the sum of average distance of the both star and mobius cube. g) Message Density(ρ): The message density is the next important measure of any network topology. It relates with the number of message sent from the source to the destination. The message density of a network should be less so that the message traffic will be minimum as the message traffic affects the communication efficiency. Theorem 7: The message density of SMQ(n, k) is represented as = . Proof: Message density is defined as =(d*N)/E ,i.e. (the average distance * total no. of nodes) / total no. of edges. From the Theorem 6, we can get the value of average distance d. Here, N=n!*2k and E=n!*2k-1(k+n-1) Hence, = d*n!2k/ n!2k-1(k+n-1) = 2d/(k+n-1). A comparative and brief account of the various topological parameters of the proposed network is worked out in Table1. Table1: Comparison of Topological Parameters Parameters

Hypercube[8]

Nodes Edges Degree Diameter

2m m 2m-1 m m

Cost

Mobius cube[1] 2m m 2m-1 M (m+2)/2 m>=4 MQm0 (m+1)/2 m>=1 MQm1 m(m+2)/2 m>=4 MQm0 m(m+1)/2 m>=1 MQm1

Average Distance(d)

Message Density

m/2

d=m/3+[1-(1/2)m]/9 for MQm0 d=m/3+[1-(1/2)m]/9 +1 for MQm1 2d/m

Star graph[4]

Star Cube[5]

SCQ[6]

SMQ [proposed]

k! k!(k-1/2) k-1 3/2(k-1)

k!2m k!2m-1 (m+k-1) (m+k-1) m+ 3/2(k-1)

k!2m k!2m-1 (m+k-1) (m+k-1)

k!2m k!2m-1 (m+k-1) (m+k-1) 3(k-1)/2+ (m+2)/2 m>=4 SMQm0 3(k-1)/2+ (m+1)/2 m>=4 SMQm1 .

(m+k-1) (m+ 3/2(k-1))

(m+k-1) (m+1/2 + 3/2(k-1))

(m+k-1)(3(k-1)/2 (m+2)/2) m>=4 SMQm0 (m+k-1)(3(k-1)/2 +(m+1)/2) m>=4 SMQm1

k4+2/k+

m/2+k4+2/k+

(11x+4y/8)+ k4+2/k+

d = m/3+[1-(- 1/2)m]/9 +k-4+2/k+ ≤ (d ) ≤ m/3+[1-(-1/2)m]/9 +1+k-4+2/k+

2d/(k-1)

2d/(m+k-1)

(k-1) 1)

3/2(k-

m+1/2 + 3/2(k-1)

IV. Broadcasting The present section is devoted for illustrating the process of Broadcasting in the proposed SMQ network topology. The parallel algorithms often require that a processor should send data to all other processesors, this is known as braodcasting. For an interconnection network, it is essential that it must broadcast messages efficiently to other nodes. There are two main situations of broadcasting: one-to-all broadcast and all-to-all broadcast. In one-to-all broadcast a single node transfers its data to all other nodes and in all-to-all broadcast every node broadcasts data to every other nodes. Theorem 7: The one-to-all broadcast algorithm for the SMQ(n, k) takes O(k + nlogn) time. Proof: In One to all broadcasting message transmits from the source node(s) of mobius cube to destination node(v) another mobius cube inside the star i.e. s=<0,123> and v=<0,213>. So message is broadcast in both the mobius cube and the star. One to all broadcasting in mobius cube takes k communication steps and star graph takes nlogn steps. As broadcast of message is done in both ways so in all the SMQ takes (k+nlogn) comuunication steps for broadcasting. Algorithm: One-to-All Broadcasting Broadcast (u, v, msg) /* u= source node v= destination node msg=message */ Step 1: Send message (msg) to neighbour node along the z-axis of u in mobius cube and one node of star (i.e. to a node of another mobius cube in star base)

Page 65

D. Pattanayak et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 62-68

Step 2: Send the msg from the received node to nodes along y-axis of the sender node and repeat step 1 for the other mobius cube. Step 3: Send the msg from the received node to nodes along x-axis and continue step 2 for other mobius cube. Step 4: Continue the step 3 for rest of nodes till all nodes of all mobius cubes receive the msg of source node. Step 5: end

Fig.4: Broadcasting in SMQ(123) and SMQ(321) Illustration: In Fig.4, we have taken two cubes of star mobius cube. The message is broadcast from the source node SMQ(123,0). In step 1the message will broadcast from (123,0) to node along z-axis i.e (123,4) and one node of star i.e. SMQ(321,0). In step 2 the message is broadcast from (123,0) and (123,4) to nodes along y-axis i.e. 0 to 2 and 4 to5 and in next mobius cube the message is broadcast from SMQ(321,0) to along z-axis SMQ(321,4) and SMQ(231,0). Similarly, step 3 is repeated. The steps are marked in bold arrow. Theorem 8: The all-to-all broadcast algorithm for the SMQ(n,k) takes O(M + nlogn) time, where M no. of k mobius cubes take part in the broadcasting. Proof: In all-to-all broadcast each node of every mobius cube transmits message to other nodes in star. i.e. each node transmits its data to all other nodes and also receives data from all nodes. 0-type mobius cube takes atmost (n+2)/2 communication steps and 1-type mobius cube executes in atmost (n+1)/2 . As all the cubes take part in broadcast at a time and if there is M number of mobius cubes, then all-to-all broadcast will take (M + nlogn) communication steps. Hence, the theorem is proved. V. Routing This section explains the process of routing in the proposed SMQ(n, k) topology. The routing is a mechanism in which the path to forward message from source to destination is determined. In routing, it needs not to visit all the nodes unlike broadcast. However it should determine the shortest path from the source to the destination. The routing algorithms can be applied to both the star graph and the mobius cube. In the SMQ(n, k) routing algorithm works in two steps: a) Routing for mobius cube b) Routing for star graph Algorithm: SMQ Routing(s, d, m) Step 1: Perform E-cube routing from the source node to other nodes inside the mobius cube. Step 2: From node of one mobius cube to node of other cube in star platform send message in the shortest path to the destination. Step 3: For each intermediate node concatenate the path from the source to the destination. Illustration: Let us assume that the message will be forwarded from the source <000,123> to the destination <001,213>. In point to point routing the is path: <000,123> <100,123> <000,213> <001,213> and the distance = 3. Hence length of the path is the sum of length inside the mobius cube and length from mobius cube of source star node to next destination star node i.e. u, v belong to mobius cube and w is at another mobius cube. Then the routing path from u to w is the sum of length of u to v and length of v to w. VI. Embedding of Networks The embedding of a network in another network is an interesting area of research. Here, we consider the embedding of Ring network and binomial tree in the propose SMQ ( n, k) topology. a) Embedding Ring A ring can be efficiently embedded in the Star mobius cube. Using gray code a network can be embedded in another network. The Figure 5 shows embedding of ring with SMQ123 and SMQ213. The function value of gray code is G (i, d) where i= node index of ring and d= dimension of SMQ.

Page 66

D. Pattanayak et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 62-68

G(0,1) = 0 , G(1,1) = 1 G(i,x+1) = The gray code of a ring with 8 nodes is G= (0,1,2,3,6,7,5,4). The 8 node ring can be embedded to 8 node i.e. 3dimensional mobius cube in this way: R(0)= <123,0> R(1)= <123,1> R(2)= <123,2> R(3)= <123,3> R(7)= <123,4> R(6)= <123,5> R(5)= <123,7> R(4)= <123,6> b) Embedding Binomial tree The Binomial tree follows the principle of divide and conquer. We can successfully map Star mobius cube into the binomial tree. The figure 6 is the partial binomial tree representation of Star mobius cube. The Binomial tree has a regular structure. The nodes of the binomial tree are notified by SMQ(3,3) addresses. Every vertex of Starmobius cube is the root of atleast one binomial tree.

Fig.5 Ring embedding in SMQ123 and SMQ213

Fig.6 Binomial tree of SMQ(3,3)

VII. Results and Discussions In this section, the results of comparison of various performance parameters of the Star-Mobius cube (SMQ) with other networks are presented. The Table 1 presents the various topological parameters of mobius cube and five related network topologies. The various candidate networks considered here for the purpose of comparison are Hypercube(HC), Star, Starcube(SC), Star crossed cube(SCC), Mobius cube(MQ). In Fig. 7, the network degree is compared with respect to other networks. As it is a product graph of the MQ and star, it is quite obvious for the Star mobius cube to have higher degree than Hypercube and Star. In other words, the proposed network Star mobius cube can connect more number of nodes than other networks. The comparison of diameter is shown in Fig 8. The Star mobius cube is observed to have lower diameter than the Star cube network. This adds to the low communication cost of SMQ and is therefore considered to be an advantage of the proposed topology. The cost of Star mobius cube is lower than Star cube as per the comparison shown in Fig 9. The Fig 10 shows that the Star mobius cube has a lower average distance than the Star cube and Star Crossed cube. The message density versus dimension is shown in Fig 11. Overally, the proposed topology Star- Mobius cube (SMQ) is observed to perform better when compared with other networks. 25

30 HC STAR SMQ

700 HC STAR SMQ SC

HC MQ STAR SC SMQ

600

500

Cost-->

Diameter-->

400

300

10 200 5

6 8 Dimension-->

Fig.7 Comparison of Degree

100

6 8 Dimension-->

Fig.8 Comparison of Diameter

6 8 Dimension-->

HC MQ STAR SC SMQ SCC

) HC MQ STAR SC SMQ SCC

1.8

1.6

Message Density-->

Fig.9 Comparison of Cost

Average Distance -->

Degree-->

20 15

1.4

1.2

0.8

6 7 Dimension-->

Fig.11 Comparison of Message density ) Page 67

D. Pattanayak et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 62-68

VIII. Conclusions In this paper, we proposed a new network topology called Star- mobius cube for large scale parallel processing. The different topological parameters of the new topology are discussed. Two algorithms one for the broadcasting and the other for routing are proposed. Embedding of the new topology with the ring and binomial tree is described The various performance features of the proposed network are analysed and compared with other cube based product graphs in terms of average distance, message density , diameter and cost. Based on comparison and analysis the Star mobius cube is found to be a better network topology in comparison to other networks. IX. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]

References

P. Cull and S.M. Larson, “The Möbius cubes”, IEEE Trans. on Computers 44 (5) (1995) 647–659. K. Efe, “The Crossed cube architecture for parallel computing”,IEEE Trans. Parallel Distributed Systems 3 (5) (1992), 513–524. T.Y. Feng,”A survey of interconnection networks”, IEEE Computer,vol.14(1981)12-27. S.B. Akers, D. Harel and B. Krishnamurthy, “The star graph: an attractive alternative to the n-cube”, International Conference on Parallel Processing (1987), 1249-126. C.R. Tripathy, “Star-cube: a new fault tolerant interconnection topology for massively parallel systems”, Journal of the Institution of Engineers (India) 84(2004), 83-92. N.Adhikari and C.R.Tripathy,”Star-crossed cube:an alternative to star graph”, Turkish Journal of Electrical Engineering & Computer Sciences (To appear) Jun- Ming XU, Meijie MA, “Survey on path and cycle embedding in some networks”, Front. Math,China 2009, 4(2): 217–252 Y. Saad and M.H. Schultz,”Topological properties of hypercubes”, IEEE Trans. on Computers 37(1988), 867-872 . Yamin Li, Shietung Peng and Wanming Chu,”Metacube– A New Interconnection Network for Large Scale Parallel Systems”, ACSAC02, Australian Computer Science Communications, 24 (4) ( 2001) 29–36. Yamin Li and S Peng,” Dual-cubes: A new interconnection network for high-performance computer clusters”, International Computer Symposium, Workshop on Computer Architecture, (2000). Hilbers, Koopman, van de Snepscheut, ” The twisted cubes”, Lecture Notes in Computer Science,Springer, (1987) 258-259. Chang Hsiung Tsai, “Embedding of meshes in Möbius cubes”, Theoretical Computer Science, Elsevier,(2008) 181-190

Page 68

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Diffraction Ring Technique and Nonlinear Optical Properties of 5Aminoindazole Abdulameer Imran, *Hussain A. Badran, Qusay Mohammed Ali Hassan Department of Physics, College of Education for pure sciences University of Basrah Basrah, Iraq Abstract: The nonlinear optical properties of 5-Aminoindazole in Dimethyl sulfoxide (DMSO) solvent was studied using single beam Z-scan technique with a continuous-wave radiation at 473 nm of an output power of 4.6 mW. All the solution samples showed large nonlinear refractive indexice and absorption coefficient of the order of 10-8 cm2/W and 10-3 cm/W, respectively. The concentration-dependent nonlinear refractive index was also investigated. We presents experimental evidences of observing diffraction pattern in 5-Aminoindazole: DMSO solvent with the calculation of the refractive index change, ∆n , the relative phase shift,  , and effective nonlinear refractive index ,n2. The solvent of spectroscopic grade and was used without further purification. All the solutions used for the study were freshly prepared. Keywords: nonlinear refraction index; Z-scan ; cw laser; diffraction rings I. Introduction Organic dyes have many advantages over other nonlinear optics (NO) materials. Photoisomerization of organic molecules enables modifies their linear and nonlinear polarizability of them as well as optical nonlinear refraction. Since the optical properties of organic molecules can be controlled optically, it has intrigued considerable interest of people [1,2]. The nonlinear optical phenomena of organic dyes can result from electronic response and/or nonelectronic one. The electronic nonlinearity is induced by either population redistribution or distortion of electronic clouds. A molecule undergoes a transition from its ground state to its excitation state after absorbing a photon. The dipole moment of the molecule changes during such a transition. The change in the dipole moment will give birth to electronic nonlinearity. A nonelectronic response is a nonradiative interaction such as cis-trans isomerism, the changes in density and temperature [3–5]. It has been well known that the nonlinear optical behavior of materials can vary greatly by changing different laser duration or different laser wavelengths. Thus, studies about the mechanism of their nonlinear optical response with different laser duration or different laser wavelengths are expected to be more interesting and important. If the nonlinear mechanism is understood for certain laser pulses, the NLO properties optimization can be well accomplished. Zscan technique is a simple and effective tool to determine the nonlinear properties [6]. It has been widely used in material characterization because it provides not only the magnitudes of the real part and imaginary part of the nonlinear susceptibility, but also the sign of the real part. Both nonlinear refraction and nonlinear absorption in solid and liquid samples can be measured easily by Z-scan technique, which use the change of transmittance of nonlinear materials [5]. In this work, we demonstrate the optical nonlinearities of a 5-Aminoindazole at different concentration in Dimethyl sulfoxide (DMSO) through Z-scan technique under laser excitation at 473 nm cw solid state laser with an output power of 4.6 mW and presents experimental evidences of observing diffraction pattern in 5Aminoindazole: DMSO solvent with the calculation of the refractive index change, ∆n , the relative phase shift,  , and effective nonlinear refractive index ,n2 . II. Materials and Methods A. Absorption spectra The solution samples of 5-Aminoindazole were prepared in DMSO. The former was contained in a 1mm quartz cuvette. The linear absorption spectrum of the sample solution with the concentrations of 2mM, 4mM, 6mM and8mM in DMSO solvent is shown in Fig.2, which was acquired using a UV–VISNIR spectrophotometer (Type: CECIL –CE-3550) . The Z-scan experiments were performed using a 473 nm solid state laser beam, which was focused by +50 mm focal length lens. The laser beam waist 0 at the focus is measured to be 22.19 μm and the Rayleigh length ZR =3.27 mm. The schematic of the experimental set up used is shown in Fig.2. A 1mm wide optical cell containing the solution of 5-Aminoindazole is translated across the focal region along the axial direction that is the direction of the propagation laser beam.

Page 69

Abdulameer Imran et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 69-74

Figure 1 UV-Visible absorption spectrum with different concentrations. Inset shows the chemical structure of 5-Aminoindazole.

The transmission of the beam through an aperture placed in the far field was measured using a photodetector fed to the digital power meter. For an open aperture Z-scan, a lens was used to collect the entire laser beam transmitted through the sample replaced the aperture. Figure 2 Schematic diagram of experimental arrangement for the Z-scan measurement.

III . Results and Discussions A.

Z-scan Measurements

The third-order nonlinear refractive index n2 and the nonlinear absorption coefficient Î˛, of the 5Aminoindazole in DMSO at various concentrations for the incident intensity I 0 = 0.594 kW/cm2 were evaluated by the measurements of Z-scan. Figure 3 shows the open Z-scan curve for the sample in solution at various concentrations. The typical Z-scan data with fully open aperture is insensitive to nonlinear refraction; therefore, the data is expected to be symmetric with respect to the focus, but absorption saturation in the sample enhances the peak and decreases the valley in the closed aperture Z-scan curve and results in distortions in the symmetry of the Z-scan curve about Z = 0 [7]. Figure 3 Open aperture Z-scan data for different concentrations.

Page 70

Abdulameer Imran et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 69-74

Figure 4 Closed-aperture Z-scan data for different concentrations.

The measurable quantity TP V can be defined as the difference between the normalized peak and valley transmittances,

TP  TV . The variation of this quantity as a function of |  0 | is given by [8]

TPV  0.406(1  S )0.25  0

…………………………………………………………..(1)

Z-scan with a fully open aperture (S=1) is insensitive to nonlinear refraction (thin sample approximation). The aperture linear transmittance is given by S  1  exp( 2ra / a ) , with ra =2.5 mm the aperture radius ,ωa = 5 mm the radius of the laser spot before the aperture,

 0 is the on-axis phase shift. The on axis phase shift is

related to the third-order nonlinear refractive index by  0  kn2 I 0 Leff ………………………………………………………………….……. (2) where k  2 /  is the wave number and sample , and



Leff  (1  exp(  0 L)) /  0 is the effective thickness of the

is the linear absorption coefficient, L the thickness of the sample,

I 0 the on-axis irradiance at focus

n 2 the third-order nonlinear refractive index. The defocusing effects of the sample in solution at various

Page 71

Abdulameer Imran et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 69-74

concentrations are shown in Fig. 4. This defocusing is attributed to a thermal nonlinearity resulting from absorption of radiation at 473 nm. Localized absorption of a tightly focused beam propagating through an absorbing dye medium produces a spatial distribution of temperature in the sample solution and, consequently, a spatial variation of the refractive index, that acts as a thermal lens resulting in severe phase distortion of the propagating beam. Generally the measurements of the normalised transmittance versus sample position, for the cases of closed and open aperture, allow determination of the nonlinear refractive index , n 2 , and the reversible saturation absorption (RSA) nonlinear coefficient,  , [9,10]. Here, since the closed aperture transmittance is

affected by the nonlinear refraction and absorption, the determination of n 2 is less straightforward from the closed aperture scans. Therefore, it is necessary to separate the effect of nonlinear refraction from that of the nonlinear absorption. A simple and approximate method [11] to obtain purely effective n 2 is to divide the closed aperture transmittance by the corresponding open aperture scans (see Figure 5) . With an open aperture the sample's transmittance is related to the nonlinear absorption coefficient through the relation [12]: 2 2T ……………………………………………………………………………(3)  I  Leff in which β is the nonlinear absorption coefficient, ΔT is one-valley transmission, The experiment was repeated for the pure solvent Dimethyl sulfoxide to account for its contribution, but no significant measurable signals were produced in either the open or the closed Z-scan traces. The nonlinear parameters calculated are as tabulated in Table 1. Figure 5 Pure nonlinear refraction curve sample in solvent at various concentrations .

Table1. Nonlinear parameters of 5-Aminoindazole solutions. Concentration ∆Φ x10 3 n2 x10 8 nx10 4 (mM) 2 4 6 8

0.36 0.59 0.74 0.94

0.27 0.45 0.56 0.73

cm 2 / W

cm / W

4.68 7.65 9.55 12.28

0.62 0.93 1.28 1.52

Page 72

Abdulameer Imran et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 69-74

B. Diffraction Ring Patterns Measurements The experimental setup for the diffraction ring patterns are the same as mentioned, expect that’s the power meter detector is replaced by the transparent screen. We can estimate the induced refractive index change, ∆n, and the effective nonlinear refractive index, n2, for the preceding data as follows. Because the laser beam used in the experiment has a Gaussian distribution, the relative phase shift,  , suffered by the beam while traversing the sample of thickness (L) can be written as [13]:

Δ  kLn

……………………………………………………………………….…………(4)

Where k=2π/λ is the wave vector in vacuum and λ is the laser beam wavelength. The on-axis nonlinear phase-shift,  , can be related to the number of rings , N , observed as [14] ……………………………..………………………….…………………..(5)   2N n can be related too to the total refractive index of the medium , n , and the back ground refractive index ,

n , as [12] …………………………………….………………..….………….……..(6) n  n  n where n0 is the background refractive index. and

n  n2 I

………………………………..……………………………………………(7)

By the combination of equations (4-7) one can calculate, ∆n ,  and n2 . As given in Table 2, for the same power (72 mM) the number of rings N for 2 mM concentration, observed is 4 while for 8 mM the number of rings N is 6. The diffraction ring patterns for the 5-Aminoindazole solutions are shown in Fig.6. Table 2. Nonlinear parameters of 5-Aminoindazole solutions using diffraction ring patterns. Rings No. 4 6

Concentration mM 2 8

n2×10-8 (cm2/W) 0.35 0.54

∆n×10-4

∆Φ

0.33 0.50

0.43 0.65

Figure 6 diffraction ring patterns for the 5-Aminoindazole solutions at 72 mW (a) 4 and (d) 6 .

V. Conclusion In summary, we have measured the nonlinear refraction index coefficient

n 2 and the nonlinear

absorption coefficient  for solutions of 5-Aminoindazole for various concentrations using the Z-scan technique with 473 nm of solid state laser (SDL). The Z-scan measurements indicated that the sample exhibited large nonlinear optical properties. We have shown that the nonlinear absorption can be attributed to a saturation absorption process, while the nonlinear refraction leads to self-defocusing in this dye. All the solutions samples showed a large nonlinear refractive index of the order of 10-8 cm2/W and 10-3 cm/W, respectively. Experimental results of ring patterns suggest the possibility of using 5-Aminoindazole for various concentrations solvent in DMSO in all optical systems. These patterns were generated in 5-Aminoindazole solution by the irradiation with

Page 73

Abdulameer Imran et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 69-74

visible laser beam of Gaussian extent .The instantaneous formation of rings prove the fast response of this substance .The thermal number of rings observed increases with increasing input power nonlinearly. The stability of the ring patterns suggest the stability of such medium. The calculation of the refractive index change, ∆n , the relative phase shift,  , and effective nonlinear refractive index ,n2 were measured too. All these experimental results show that the solutions of 5-Aminoindazole are a promising material for applications in nonlinear optical devices. References [1] [2] [3] [4] [5] [6] [7]

[8] [9]

[10] [11] [12]

[13] [14]

P. Wang, H.Ming, J. Zhang, Z. Liang, Y.Lu, Q. Zhang, J. Xie and Y. Tian, Nonlinear optical and optical-limiting properties of Azobenzene liquid crystal polymer Opt. Commun. Vol.203(1-2), March 2002. pp.159-162. doi:10.1016/S0030-4018(02)01098-2 J.C. Liang and X.Q. Zhou," Application of continuous-wave laser Z-scan technique to photoisomerization ", J. Opt. Soc. Am. B, vol.22(11),November 2005, pp.2468-2471. doi:10.1364/JOSAB.22.002468 R.Rangel Rojo, S.Yamada, H.Matsuda and D.Yankelevich,''Large near-resonance third-order nonlinearity in an azobenzenefunctionalized polymer film '' Appl. Phys. Lett. vol.72(9) , 1998, pp.1021-1023.doi: 10.1063/1.120977 . C. Gayathri, and A. Ramalingam, " Investigation of optical nonlinearities of an azo dye using a 532 nm diode-pumped Nd:YAG laser Spectrochim " Acta Part A vol.69(1), January 2008 , pp.96-101. doi.10.1016/j.saa.2007.03.025. T. ChaoHe, C. SunWang, "Study on the nonlinear optical properties of three azo dyes by Z-scan measurements", J. Mod. Opt. Vol.55(18), December 2008, pp.3013-3020. doi:10.1080/09500340802296307. J.Ting Jian, S. Zhiguo and C.Yong Guang,''The nonlinear optical response of a fluorine-containing azoic dye'', Opt. Commu. Vol.283(6), 15 march, 2010, pp.1110-1113 . doi.10.1016/j.optcom.2009.10.110. C. Zhang, S.Ying Lin, W.Xin , E.Fritz Kühn, W.Yu Xiao, X.Yan and X..Xin Quan."Large third-order optical nonlinearity of two cubane-like clusters containing oxotrithiometalate anions and silver: synthesis, characterization, reactivity, and NLO properties– structure correlation" J.Mater.Chem, vol.13, January 2003, pp.571-579. doi: 10.1039/B205450G M. A. Quasy. and P.K. Palanisamy,Investigation of nonlinear optical properties of organic dye by z-scan technique using He–Ne laser Optik, vol.116, 2005, pp.515–520. doi.10.1016/j.ijleo.2005.05.001 S.J.Mathews, S. Chaitanya Kumar, L. Giribabu and S.Venugopal Rao, " Nonlinear optical and optical limiting properties of phthalocyanines in solution and thin films of PMMA at 633 nm studied using a cw laser "Mater. Lett. vol.61(22), Septemper 2007, pp.4426-4431. doi.10.1016/j.matlet.2007.02.034. H. A.Badran, Q. M. A. Hassan, A. Y.Al-Ahmad, C. A.Emshary," Laser-induced optical nonlinearities in Orange G dye: polyacrylamide gel " Can. J. Phys. Vol.89, November 2011, pp.1219-1224. doi:10.1139/P11-118. M. Sheik-Bahae, A.A. Said, T.H.Wei, D. J. Hagan and E.W.Van Stryland, “Sensitive measurement of optical nonlinearities using a single beam,” IEEE. J. Quant. Elect. vol.26, 1990, pp.760-769. doi:: 10.1109/3.53394. K.Milanchian, H.Tajalli, A.Ghanadzadeh Gilani and M.S. Zakerhamidia, “Nonlinear optical properties of two oxazine dyes in aqueous solution and polyacrylamide hydrogel using single beam Z-scan,” Opt. Mat. vol. 32(1), November 2009, pp.12–17. doi:10.1016/j.optmat.2009.05.011 K.Ogusu, Y. Kohtani and H.Shao, “Laser-induced diffraction rings from an absorbing solution,” Opt. Rev. vol.3, 1996 , pp.232–234. doi: 10.1007/s10043-996-0232-1 A.B.Villafranca and K. Saravanamuttu, “ Diffraction rings due to spatial self- phase modulation in a photopolymerizable medium,” J. Opt. A: Pure Appl. Opt. vol.11, 2009, pp. 125202. doi:10.1088/1464-4258/11/12/125202.

Page 74

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Seismic Data Analysis in Odyssey Software 1

Nazarov Yuri P., 2Poznyak Elena V., 3Filimonov Anton V., 1 Central Research Institute of Building Constructions 2 National Research University Moscow Power Engineering Institute 3 Institute of Computer-Aided Design RAS Moscow, RUSSIA. Abstract: This article describes tools of Odyssey software (Eurosoft Co., Russia), that help engineers to create more accurate and reliable integrated seismic action model and make seismic data analysis. The main instruments are based on vector, correlation and spectral analysis of the seismic data. For time-domain analysis initial accelerograms are scaled, changed the range and filtered. Filtering based on Fourier transform with removing shortest waves that are safe for the construction. It is possible to create a generalized seismic wave model to compute rotations of a building. For frequency-domain analysis by modal response spectral method the amplification factors, response spectrums and theirs envelopes are calculated for a translational and, if it is necessary, rotational motion. Keywords: seismic analysis; seismic rotations; seismic stability; response spectral method; seismic action; accelerograms I.

Introduction

The Odyssey software is developed by Eurosoft Co. (Moscow, Russia). This is a handy tool to calculate initial data for integrated model of seismic action. Theoretical fundamentals and algorithms are described in [1-3] and based on multi-year experience in the field of seismic stability analysis in Central Research Institute of Building Constructions named after V.A. Kucherenko, Moscow, Russia. The main features of software are describes in this paper (hereinafter options of the menu item Calculation (Fig.1) are italic). Figure 1. Odyssey software. Accelerograms of natural earthquake intended for seismic analysis.

II.

Calculations

Accelerograms for seismic analysis is shown in Figure 1. The menu Calculation is shown in Figure 2. Menu Calculation includes scaling, filtering waves in frequency and wavelength, changing of time samples, vector, correlation, spectral analysis and calculating of amplification factors. Scaling of accelerograms is used when it’s necessary to switch to other dimensions of physical quantities, or if there is a scaling coefficient device during measurements. The filtering allows excluding high frequency components or wavelengths, which are much smaller than structure’s dimensions in the plan, from the spectrum of the seismic action. These components are not dangerous for buildings. The changing of the time samples is an important procedure to obtain data for further seismic computing. This procedure allows to remove insignificant P-waves (primary waves or compression wave) and consider only the intensive part of S-waves (secondary or shear). In addition, removing of the low intensity parts reduces the overall

Page 75

Nazarov Yuri P. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 75-77

nonlinearity of the seismic action. This instrument helps to select dangerous part of random process and, considering it as stationary, to perform a further correlation and spectral analysis more correctly. Input data for seismic analysis are natural or synthetic accelerograms of translational ground motion in three orthogonal directions. Such data allows representing the seismic action in the form of three-dimensional random vector with time-depending components. Figure 2. Tools of Calculation item

There are operations for vector analysis: coordinate system rotation, calculation of the seismic vector module and direction cosines, calculation of rotational components. The coordinate system rotation is needed to direct the global axes of the structure with axes that specify seismic wave impact. The module of the seismic action is invariant parameter (independent of the chosen coordinate system), characterizes the vector magnitude of seismic acceleration. The rotational components of seismic ground motion are components of the angular accelerations vector relative to the corresponding axes. To compute rotations of a building we should determine a generalized seismic wave model [1-4]. Examples of accelerograms with corresponding rotations are shown on Figures 3-6. Figure 3. Synthetic accelerogram Figure 4. Natural accelerogram (N-S direction) (N-S direction)

Figure 5. Rotation for synthetic accelerogram (about Z-direction)

Figure 6. Rotation for natural accelerogram (about Z- direction)

The software has tools for spectral and correlation analysis to determine required statistical characteristics of random processes. The correlation analysis includes the calculation of the normalized autocorrelation and cross-correlation functions of random processes of translational or rotational motions relative to the given direction. Expectation value, standard and correlation matrix are determined for the following random processes: three components of the acceleration vector of translational motion, the module of this vector and three direction cosines. The obtained parameters of random processes are first and second order moments of seismic action. These data may be used to determine a buildingâ&#x20AC;&#x2122;s response by using spectral representations method according to the stochastic formulation [5].

Page 76

Nazarov Yuri P. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 75-77

The spectral analysis of random processes includes direct and inverse Fourier transforms of seismic vector components (including rotational components), the dependence of the standard of seismic action on frequency and spectral power density. There are six degrees of freedom of ground volume motion for integrated seismic model: first three are translational and the other ones are rotational. According to the conventional quasi-static deterministic approach [6, 7], the horizontal and vertical seismic loads are the product of translational acceleration on the appropriate nodal mass. Obtained products are multiplied by amplification factors to account the dynamic effects. If the seismic loads are random, the structural response is a random process too. Amplification factor is defined as the ratio of standard of generalized coordinate to the value of this coordinate in the case of static action. Odyssey software calculates the amplification factors for the translational motion in a given direction as a function of natural periods or frequencies. To obtain a response spectrum the amplification factor should multiply on the standard of corresponding accelerogram. Furthermore, there is the possibility to draw envelopes for amplification factors for each or all directions. The example of amplification factors and their envelope are shown in Figure 7. In some cases [1-3, 5], there are additional loads from seismic rotations. Odyssey computes angular accelerations of the rotational movement (rotational components) as a function of time and amplification factors for them. Thus, software provides data for integrated model of seismic actions. The amplification factors and their envelopes are data for quasi-static spectral analysis. The accelerations of translational and rotational motions are data for verification analysis in the time domain. All seismic data may be exported to different engineering software by using simple text format files and ready to use in FEM software STARK ES [8, 9]. Figure 7. Amplification factors and envelope for translational motion

III. [1] [2] [3] [4] [5] [6] [7] [8] [9]

References

Nazarov Y.P. The analytical calculation fundamentals of constructions on seismic actions. -М.: Science, 2010. Nazarov Y.P. Models of seismic actions for calculation. -М.: Science, 2012. NikolaenkoN.A., NazarovY.P. Dynamics and seismic resistance of structures. -М.: Stroyizdat, 1988. N.M.Newmark, E. Rosenblueth. Fundamentals of Earthquake Engineering. -М.: Stroyizdat, 1980. Bolotin V.V. Statistical Methods in Structural Mechanics.-M. Stroyizdat, 1965.-279 p.p. SP 31-114-2004. RULES FOR DESIGN OF HOUSES AND PUBLIC BUILDINGS TO BE CONSTRUCTED IN SEISMIC REGIONS.- М.: 2005. SNIP II-7-81 *Building Regulations. Construction in seismic regions. Moscow, 2002. STARK_ES. User’s guide. Structural analysis software. Strength, stability and vibration of building structures. ©Eurosoft Co., Moscow, 2006. EUROSOFT Odyssey. User’s guide. ©Eurosoft Co., Moscow, 2013

Page 77

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net KINETICS AND MECHANISM OF ZnCl2 CATALYSED OXIDATION OF ORGANIC SULFOXIDES BY PERMANGANATE IN NONAQUEOUS MEDIUM K.P.Srivastava* & Sanjay Kumar Rai P.G. Department of Chemistry Ganga Singh College, Jai Prakash University, Chapra-841 301 Bihar, INDIA _______________________________________________________________________________________ Abstract: The rates and mechanism of zinc chloride (Lewis acid) catalysed oxidation of several organic sulfoxides by permanganate have been studied in anhydrous acetone solutions. The kinetic rate law obtained indicates that a complex between permanganate and the Lewis acid (ZnCl 2) is formed before oxidation of the sulfoxides occurs. The function of the zinc chloride is to enhance the reactivity of permanganate. A Hammett ρ value of -2.1 8 ± 0.09 is obtained for the oxidation of sulfoxides at 23.0 0C. A negative ρ value means that the sulfur is electron deficient in the transition state. Values for the HOMO energies of sulfoxides and the LUMO energies for permanganate ion and the permanganate-zinc chloride complex have been calculated. The results indicate that electron donating substituents on the ring increase the rate of the reaction and electron withdrawing groups slow down the reaction. Keywords: Sulfoxides, permanganate, acid catalysed oxidation, LUMO, HOMO. _________________________________________________________________________________________ I. INTRODUCTTION Many sulfur-containing compounds are present in natural biological systems and play key roles in the activity of some enzymes [1-2]. An understanding of the deactivation of these enzymes by oxidation requires knowledge of the mechanisms by which sulfur compounds are oxidized [3].The observation that Cytochrome P-450 can readily catalyze the oxidation of nitrogen and sulfur compounds [4] relates a study of the reactivity of transition metal-oxo compounds to some biological processes. For example, the transfer of oxygen from the lungs, via hemoglobin and myoglobin, to the cells where metabolism occurs involves the use of a transition metal; iron, as an oxygen transfer agent. An improved understanding of transition metal based oxygen transfer reactions may also find application in organic synthesis [5]. Oxygen transfers are necessary steps in the manufacture of a number of commercially important organic compounds. In addition, similar reactions rnay be used for environmental clean-ups that involve destruction of odoriferous sulfur containing compounds [6]. A better understanding of these reactions, therefore, has implications for both commerce and health. In continuation of research works on Lewis acid catalysed oxidation of organic-sulfur compounds [7], here we report the Lewis acid (ZnCl2) catalysed oxidation of sulfoxides by permanganate. The objective of the research paper reported herein is to improve our understanding of kinetics and mechanism of the reaction between potassium permanganate and sulfur-containing organic compounds vis. sulfoxides, when ZnCl2, a Lewis acid is present as a catalyst in anhydrous acetone as non-aqueous medium. I. MATERIALS & METHODS Potassium permanganate was BDH Analar grade. Acetone, used as the solvent in all kinetic experiments, was Fisher HPLC grade. Anhydrous zinc chloride was obtained from Aldrich and stored in a desiccator. Solvents and reagents used in synthesis were all analytical grade and used without further purification. Silica gel used in column chromatography (200 meshes) was from BDH. A. Synthesis of Sulfoxides Methyl p-nitrophenyl sulfide (22.0g, 0.13 mol) was dissolved in 125 mL acetone. Then 30% H2O2 (20 g) was added. The mixture was thoroughly stirred at room temperature until TLC showed no sulfide. The reaction mixture was extracted with CHCl3 (450 ml). The CHCl3, extracts were combined and dried overnight with MgSO4. Removal of CHCl3 from the solution by evaporation afforded a yellow solid. The crude product was a mixture of sulfoxide and sulfone which were separated by column chromatography (eluent: chloroform/ethyl acetate 1/1). Yield: 4.82 g, (20%). mp 148.50C (lit. 148-149 0C [8]). p-Methoxyphenyl methyl sulfoxide, methyl phenyl sulfoxide, p-fluorophenyl methyl sulfoxide, p-chlorophenyl methyl sulfoxide and p-nitrophenyl methyl sulfoxide were prepared in essentially the same manner except that

Page 78

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

the reaction took place at a lower temperature (using an ice-water bath). The yields, the observed melting points and structural data are summarized in Tables-1 and -2 respectively. II. KINETIC METHODS The kinetic study of the reduction of permanganate by phenyl sulfoxide and various ring substituted sulfoxides was completed using anhydrous acetone solutions containing different concentrations of ZnCl 2 as the Lewis acid. All kinetic studies were carried out under pseudo first order conditions with the substrate concentration at least ten times that of permanganate. All substrates were carefully purified before being used in kinetic studies by recrystallization, distillation or preparative chromatography. The reaction rates were determined by monitoring UV/Vis spectral changes as the reaction progressed. The decrease in absorbance at 550 nm was followed on the spectrophotometer. In the experiment, a solution of zinc chloride in anhydrous acetone was sealed in a 50 mL Erlenmeyer flask and immersed in a constant temperature bath for one hour. While the zinc chloride solution was being thermostated, a permanganate solution was prepared by placing a few milligrams of KMnO 4 in a 50 mL Erlenmeyer flask and adding 40 mL of anhydrous acetone. After swirling for about 30 seconds, the supernant was transferred, using a disposable pipette, to another flask suspended in the constant temperature bath. The flask was stoppered and sealed to prevent contact with moisture in the air. An aliquot of the zinc chloride solution (2.0 mL) was then transferred to a 10 mm cuvette and a stock solution of sulfoxide in anhydrous acetone (0.10 ml) was added using a microliter syringe. The cuvette was placed in the thermostated cell compartment of the spectrophotometer, the background of the instrument was recorded, and after a few minutes, the reaction was initiated by adding permanganate solution (0.50 mL) using a microliter syringe. The cuvette was quickly inverted several times to ensure good mixing and spectra were collected every five seconds until the reaction was complete. For determination of the order with respect to zinc chloride, the concentration of zinc chloride was varied over a larger range. The concentration of sulfoxides was varied from 6.0×10 -3 to 1.3×10-1 M. Table-1: Yields, melting points of para-substituted phenyl methyl sulfoxides Substrates

Yield

MP (0C)

p-MeOPhSOCH3

42-43

p-MePhSOCH3

41-43

PhSOCH3

146-148 125-127

p-FPhSOCH3

p-ClPhSOCH3

46-47

p-NO2PhSOCH3

147-148

Table-2: NMR and IR Data for para-substituted phenyl methyl sulfoxides Substrates p-MeOPhSOCH3

H NMR (δ, ppm) 2.67(s,3H), 3.95(s,3H), 7.0(d,2H), 7.8(d,2H)

p-MePhSOCH3

2.38(s,3H), 2.65(s,3H), 7.28(d,2H), 7.68(d,2H),

PhSOCH3

2.70(s,3H), 7.50(m,3H), 7.65(d,2H)

p-FPhSOCH3

2.65(s,3H), 7.15(d,2H), 7.60(d,2H)

p-ClPhSOCH3

2.70(s,3H), 7.50(d,2H), 7.60(d,2H)

p-NO2PhSOCH3

2.70(s,3H), 7.60(d,2H), 8.35(d,2H)

IR (ʋ, cm-1) 1048(s), 1255(s), 1497(s), 1595(s), 2837-3200(s) 1056(s), 1448(m) 3000(s) 1048(s), 1481(s), 1582(s), 2900(s), 3010(s) 1049(s), 1087(s), 1493(s), 1590(m), 2900- 3100(w), 1051(s), 1476(s), 1576(m), 2359(m), 3274(m) 852(s), 1048(s), 1340(s), 1529(s), 1460(s)(nujol)

III. RESULTS & DISCUSSION The general rate law for the reaction between permanganate and sulfoxide can be expressed as in equation-1 (LA is an acronym for Lewis acid): Rate = k [MnO4-]x [RSOR’]y [LA]z (1) The orders with respect to permanganate, sulfoxide and Lewis acid are designated as x, y and z respectively. The experimentally determined rate law (first order in oxidant, first order in Lewis acid and of variable order between unity and zero in reductant) is of the form (equation-2) expected if zinc chloride acts as a catalyst by combining with permanganate prior to the redox step. Thus, the Lewis acid forms complex with the permanganate in first step of the reaction and the sulfoxides are subsequently oxidized by this complex in a second step. The corresponding rate law is given by equation -2: -d[MnO4-] / dt = k1k2[MnO4-][ZnCl2][RSOR’] / k1 + k2[RSOR’] (2) The rate constants for the oxidation of methyl phenyl sulfoxide at various temperatures, obtained using the method described for sulfides, are summarized in table-3. The oxidations of p-chlorophenyl methyl sulfoxide IJETCAS 14-116; © 2014, IJETCAS All Rights Reserved

Page 79

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

and methyl p-nitrophenyl sulfoxide are very slow. Although the plots (Ri-Rb)/Ai vs. the concentration of sulfoxide fit equation-3. The data, therefore, were found to fit equation -3 better. Under such conditions the reaction would appear to be first order with respect to sulfoxide, as indicated in equation-3: (Ri-Rb) / Ai = k1k2[ZnCl2][RSOR’] / k1 (3) Use of the same definitions as in previous sections, for K and kobs (K = k1/k2 and kobs = Kk1) gives equations 4-6 as: (Ri-Rb) / Ai = kobs[ZnCl2][RSOR’] (4) (Ri-Rb) / Ai = k’ [RSOR’] (5) K’ = kobs[ZnCl2] (6) The values for K were obtained from the slope of plots of (R i-Rb) /Ai vs the concentration of sulfoxide and values of kobs were calculated from the relationship show in equation-7: kobs = K’/ [ZnCl2] (7) The values for K’ and kobs, at 23.00C obtained in this way for investigaed sulfoxides are presented in table-3. Table-3: Rate constants for the oxidation of sulfoxides when ZnCl2, is used as Lewis acid catalyst at 230C and [ZnCl2] =1.16×10-3M Substrates

σ+

k1(M-1s-1)

Kobs(M-2s-1)

p-MeOPhSOCH3

-0.268

-0.778

337

p-MePhSOCH3

-0.170

-0.311

148

PhSOCH3

0.000

98 52

p-FPhSOCH3

0.062

0.073

p-ClPhSOCH3

0.227

0.114

p-NO2PhSOCH3

0.778

0.790

1.5

A. Role of Lewis Acid in the Reaction Lewis acids have a significant catalytic effect on the rates of oxidation of sulfoxides by permanganate. There is almost no reaction between permanganate and sulfoxide in the absence of a Lewis acid. The reaction between a Lewis acid and permanganate is consistent with the possibility that it makes permanganate a more reactive oxidant. Reaction of the Lewis acid with permanganate would cause the metal center (manganese) to become more positive, as in scheme-1. The Lewis acid acts in much the same way as a Bronsted acid does in the reactions of permanganate in protic media [9].

Scheme-1: The reaction between permanganate and zinc chloride B. Substituent Effects The rate constants for the reaction between permanganate and different aryl methyl sulfoxides at 23.0 0C are summarized in table-3. The substituted results indicate that sulfoxides with electron donating groups react faster than those with electron withdrawing groups. The Hammett plot using a substituent constants, reproduced in figure-1, gives ρ value of -2.1 8 ± 0.09 at 23.0 0C. As indicated by figure-2, the correlation when σ+ values are used is less exact. The reactions of permanganate with sulfoxides are, therefore, much more sensitive to substituent effects than the corresponding reactions of sulfides where the ρ value was observed to be only -1.09 ± 0.05 at the same temperature.

Figure-1: Hammett plot for the oxidation of methyl p-substituted phenyl sulfoxides by permanganate at 23.0 0C; {[ZnCl2] = 1. 16×10-3 M; Slope = -2.18 ±.0.09; Intercept=1.89±.03; r2 = 0.993}

Page 80

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

Figure-2: Hammett plot using σ+ values for the oxidation of methyl p-substituted phenyl sulfoxides by permanganate at 23.00C; {[ZnCl2] = 1.16×10-3 M; Slope = -1.52 ± 0.23; Intercept = 1.59 ± 0.11; r2 = 0.919} C. Activation Parameters The enthalpies and entropies of activation were determined for each substituted phenyl methyl sulfoxide using ZnC12 as the Lewis acid catalyst. The Gibbs free energies of activation for these reactions were calculated using equation-8. The data are summarized in tables-4 and -5. ∆G* = ∆H* - T∆S* (8) The activation energies, ∆G*, for these reactions were also correlated well with the calculated HOMO energies for the sulfoxides (figure-3). Because oxygen has a higher electronegativity than sulfur, the double bond between sulfur and oxygen in sulfoxides is polarized, placing positive charge on sulfur. Our theoretical calculations indicate that a charge of +1.0 resides on sulfur and - 0.8 charges on oxygen. This results in a lower electron density on sulfur in the sulfoxide than in the corresponding sulfide and decreases the ease of electron loss from sulfur. According to our theoretical calculations, the lone pair of electrons on the sulfur of a sulfoxide resides on the HOMO orbital. The HOMO orbital is primarily an oxygen p orbital. The reaction of a sulfoxide with permanganate would therefore involve an interaction between the O and Mn atoms. The HOMO energy of a sulfoxide is lower than that of the corresponding sulfide. Therefore the energy gap between the HOMO of the sulfoxide and the LUMO of the MnO 4.ZnCl2- complex is larger than that between the HOMO of the corresponding sulfide and the LUMO of the MnO 4.ZnCl2- complex [7]. This difference in energy is consistent with the difference between reaction rates of sulfoxides and sulfides and is consistent with the proposed mechanism which involves nucleophilic attack by the reductant on the oxidant. The activation energies, ΔG*, for these reactions also correlate well with the calculated HOMO energies for the sulfoxides (Figure -3).

Figure-3: Plot of HOMO energy vs. free energy of activation for ZnCl2 catalysed oxidation of sulfoxides by permanganate in non-aqueous medium

Page 81

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

Table-4: Activation parameters for the formation of the MnO4-ZnCl2 complex (k1) when sulfoxides are used as reductants; {[ZnCl2] = 1.16×10-3 M} Substrates ∆H1* (kJmol-1) ∆S1*(Jmol-1K-1) ∆G1*(kJmol-1) p-MeOPhSOCH3 59.5 -17.9 64.8 p-MePhSOCH3 59.2 -19.6 65.0 PhSOCH3 60.0 -16.4 64.9 p-FPhSOCH3 59.8 -17.4 65.0 Table-5: Activation parameters for the reaction of the MnO4 –ZnCl2 complex (kobs) with sulfoxides {∆G* is calculated at 23.00C and [ZnCl2] = 1.16 × 10-3 M} Substrates ∆H1* (kJmol-1) ∆S1*(Jmol-1K-1) ∆G1*(kJmol-1) p-MeOPhSOCH3 19.8 -132 58.9 p-MePhSOCH3 22.9 -128 60.8 PhSOCH3 24.4 -126 61.7 p-FPhSOCH3 26.3 -125 63.3 D. Linear Free Energy Relationships The activation energy of a redox reaction involving oxygen transfer reactions should be proportional to the difference between the energies of the reductants' HOMOs and the oxidants' LUMOs; i.e., the primary process involves a transfer of electrons from the reductant to the oxidant. The ΔG* values are related to the rate constants through the Eyring equation (equation-9) lnkobs =X (EHOMO / RT) + constant (9) The relationship expressed in equation-9 indicates that, under constant conditions, the natural logarithms of the rate constants should be linearly related to the energies of the highest occupied molecular orbitals of the reductants for oxidation by a particular oxidant. HOMO energies for a number of sulfoxides are presented in table-6. HOMO and LUMO energies and the interna1 energies (ΔE) for MnO 4; the Lewis acids and their complexes were calculated using Gaussian 94 with the LANL2DZ basis set [10]. Table-6: Highest Occupied Molecular Orbital Energies calculated using Gaussian 94 at the RHF/6-31G(d) level at 296.15K Compound

EHOMO (Hartree)

EHOMO (kJ/mol)

EHOMO /RT

p-MeOPhSOCH3 p-MePhSOCH3 PhSOCH3 p-FPhSOCH3 p-ClPhSOCH3 p-NO2PhSOCH3

-0.31601 -0.32790 -0.33671 -0.33996 -0.34103 -0.36114

-829.5 -860.7 -883.9 -892.4 -895.2 -948.0

-336.9 -349.6 -359.0 -362.4 -363.6 -385.0

The plot of lnkobs vs. EHOMO / RT reproduced in figure-4 has an acceptable correlation coefficient of 0.936. The two points that cause the most deviation, p-chlorophenyl methyl sulfoxide and p-nitrophenyl methyl sulfoxide, are for the two rate constants that contain the greatest experimental uncertainty. If these two points are deleted from the plot, the correlation coefficient is excellent.

Figure-4: Plot of lnkobs vs. EHOMO / RT for para-substituted sulfoxides at 296.5K [ZnCl2] = 1.16×10-3M, Slope = -0.0818, Intercept = 33.2, r2 = 0.936 The empirical observation expressed in figure-3 is consistent with Frontier Molecular Orbital Theory which suggests that activation energies should be proportional to the difference between the HOMO energy of the electron pair donor and the LUMO energy of the electron pair acceptor [11].

Page 82

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

Frontier Molecular Orbital Theory may also be used to account for the catalytic effect of Lewis acids. The calculations indicate that complexation of permanganate with zinc chloride lowers the LUMO energy by 0.06107 Hartrees (160.3 kJ/mol). Further, with reference to figure-5 it can be seen that a reduction in the energy of the oxidant's LUMO would reduce the LUMO-HOMO energy difference. If, as is predicted by this theory, the activation energy for a reaction is proportional to the LUMO-HOMO energy difference, a decrease in the energy of the LUMO or an increase in the energy of the HOMO would result in a decrease in activation energy and an increase in rate. It can be seen, therefore, that Frontier Molecular Orbital Theory accounts qualitatively for both the effect of substituents on the rate of the reaction and on the catalytic effect of Lewis acids (figure-5).

Figure-5: The energy difference between the HOMO of para-substituted sulfides and the LUMO of the MnO4.ZnCl2- complex E. Structures ZnCl2, solvated by two acetone molecules, has a tetrahedral geometry as shown in structure (figure-6). This structure was modeled theoretically. The bond lengths and angles of the theoretically optimized structure are: Bond length (in Ǻ) Bond angle (in 0) Zn-Cl = 2.307 Cl-Zn-Cl = 126.60 Zn-O = 2.043 O-Zn-O = 104.60 Cl-Zn-O = 105.10

Figure-6: Optirnized structure of ZnCl2 in acetone The structure for the permanganate-zinc chloride complex modeled is shown as structure in figure-7.The calculated bond lengths and angles in this structure are: Bond length (in Ǻ) Bond angle (in 0) Mn-O(2) = 1.592 Mn-O(2)-Zn = 152.90 Zn-O(2) = 2.054 O(2)-Zn–O(4) = 88.30 Zn-O(4) = 2.110 Cl(5)-Zn-Cl(6) = 124.90 Zn-Cl(5) = 2.231 O(4)-Zn-Cl(5) = O(4)-Zn-Cl(6) = 106.20 Zn-Cl(6) = 2.231 Zn-O(4)-C = 140.7 (4)-C(7) = 1.229 O(2)-Zn-Cl(5) = O(2)-Zn-Cl(6) = 112.20 and the rest of Mn-O bond lengths in permanganate ion are 1.556 A0, and Mn is in tetrahedral geometry.

Figure-7: Optimized structure of the MnO4.ZnCl2- complex

Page 83

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

F. The Reaction Mechanism The experimental data indicate that the equilibrium for the formation of the permanganate-zinc chloride complex is established before oxidation of the sulfide occurs (scheme-2):

Scheme-2: The equilibrium for the formation of the MnO4.ZnCl2- complex A mechanism for sulfoxide oxidation that is consistent with our experimental results and theoretical calculations is proposed in scheme-3. The formation of permanganate - zinc chloride complex is followed by a reaction in which the sulfoxide becomes a ligand on manganese. Since the highest occupied molecular orbital in the sulfoxide is oxygen 2p orbital containing a lone pair of electrons, bonding to the metal would be through oxygen rather than sulfur. The resulting positive charge would be delocalized over oxygen and sulfur as indicated by the resonance structures 1 and 2. The reaction of the electron deficient sulfur with 3, an adjacent oxide ligand would give an intermediate, containing a four membered ring, similar to a proposal previously made by Block [12]. The decomposition of 3 by transfer of electrons as indicated would give the product, a sulfone, along with manganese (V) compound which would be rapidly reduced to MnO2 under these conditions. The transition state is more organized than the ground state as suggested by the negative entropies of activation. Scheme-3: Proposed mechanism for the zinc chloride catalysed oxidation of sulfoxide by permanganate in acetone medium

IV. CONCLUSIONS The rates of ZnCl2 catalysed oxidation of several para-substituted sulfoxides by permanganate have been studied in anhydrous acetone solutions. The kinetic rate law obtained indicates that a complex between permanganate and the Lewis acid (ZnCl2) is formed before oxidation of the sulfoxides occurs. The function of the zinc chloride is to enhance the reactivity of permanganate. A Hammett ρ value of -2.1 8 ± 0.09 is obtained for the oxidation of sulfoxides at 23.0 0C. Values for the HOMO energies of sulfoxides and the LUMO energies for permanganate ion and the permanganate-zinc chloride complex have been calculated. A new linear free energy relationship, based on frontier molecular orbital theory, has been derived from first principles. This equation was used to simultaneously correlate the rate constants for the oxidation of sulfoxides with the energies of their highest occupied molecular orbital. The results indicate that electron donating substituents on the ring increase the rate of the reaction and electron withdrawing groups slow down the reaction. ACKNOWLEDGEMENT The authors are thankful to Prof. S.N.Vidyarthi, Dean, Faculty of Science, Jai Prakash University, Chapra84301, Bihar, INDIA for his kind cooperation and support.

Page 84

K.P.Srivastava et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 78-85

REFERENCES [1]. [2]. [3]. [4]. [5]. [6]. [7]. [8]. [9]. [10]. [11]. [12].

S. Oae, Organic Chemistry of Sulfur, Plenum Press, New York, 1977. E. Block, Reactions of Organosulfur Compounds, Academic Press, NewYork, 1978. C. Chen, C. S. Foote and C. L. Gu, J. Am. Chem. Soc. 1992,114. 3015 and references cited therein. R. Sato and Y. Omuru, Cytochrome P-450, Kodansha, Tokyo and Academic Press, New York 1978. A. J. Fatiadi. Synthesis, 1987, 85. Y. A. Attia and W. Lei, Processing and Utilization of High Sulfur Coals II (Eds: Y.P. Chugh and R. D. Caudle), Elsevier Science Publishing Company Inc. 1987. K. P. Srivastava and S. K. Rai, Chem. Sci. Trans., 2014, 3, (accepted). A. Cemiani and G. Modena, Gaz. Chim. Ital. 1953,119, 843. L. Levai and G. Fodor, O. Fuchs, Ber. 1960, 93, 387. S. Hu, D. M. Thompson, P. 0. Ikekwere, R. J. Barton, K. E. Johnson, B. E.Robertson, Inorg. Chem. 1989, 28, 4552. I. Fleming, Frontier Orbitals and Organic Chemical Reactions, Wiley, London, 1978. E. Block, Reactions of Organosulfur Compounds, Academic Press, New York, 1978.

Page 85

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A STUDY ON IMAGE DENOISING FOR LUNG CT SCAN IMAGES S.Sivakumar and Dr.C.Chandrasekar Department of Computer Science, Periyar University, Salem, Tamilnadu-636011, India. Abstract: Medical imaging is the technique and process used to create images of the human body for clinical purposes and diagnosis. Medical imaging is often perceived to designate the set of techniques that noninvasively produce images of the internal aspect of the body. The x-ray computed tomographic (CT) scanner has made it possible to detect the presence of lesions of very low contrast. The noise in the reconstructed CT images is significantly reduced through the use of efficient x-ray detectors and electronic processing. The CT reconstruction technique almost completely eliminates the superposition of anatomic structures, leading to a reduction of "structural" noise. It is the random noise in a CT image that ultimately limits the ability of the radiologist to discriminate between two regions of different density. Because of its unpredictable nature, such noise cannot be completely eliminated from the image and will always lead to some uncertainty in the interpretation of the image. The noise present in the images may appear as additive or multiplicative components and the main purpose of denoising is to remove these noisy components while preserving the important signal as much as possible. In this paper we analyzed the denoising filters such as Mean, Median, Midpoint, Wiener filters and the three more modified filter approaches for the Lung CT scan images to remove the noise present in the images and compared by the quality parameters. Keywords: Medical CT scan images; Noise removal; Statistical filters; Quality measures I. Introduction Medical image enhancement technologies have attracted much attention since advanced medical equipments were put into use in the medical field. Enhanced medical images are desired by a surgeon to assist diagnosis and interpretation because medical image qualities are often deteriorated by noise and other data acquisition devices, illumination conditions, etc. Our targets of medical image enhancement are mainly to solve problems of the high level noise of a medical image. The noise present in the images may appear as additive or multiplicative components and the main purpose of denoising is to remove these noisy components while preserving the important signal as much as possible[1]. The Medical Images normally have a problem of high level components of noises. There are different techniques for producing medical images such as Magnetic Resonance Imaging (MRI), X-ray, Computed Tomography and Ultrasound, during this process noise is added that decreases the image quality and image analysis. Image denoising is an important task in image processing. II. Image Noise There are several types of image "noise" that can interfere with the interpretation of an image. Although noise may infiltrate and corrupt the data at any point in the CT process, the ultimate source of noise is the random, statistical noise, arising from the detection of a finite number of x-ray quanta in the projection measurements. Properties of CT Noise: The consequences of statistical noise in CT reconstructions have been discussed by numerous authors [1][2][3]. Several of these authors have pointed out that the process of reconstruction leads to some peculiar characteristics of the noise in CT images. The properties of statistical (quantum) noise in CT reconstructions will be explored in this discussion. Although the precise random noise pattern of any image cannot be predicted a priori, it is possible to characterize the average behavior of the noise by a variety of methods. Some of these methods give a complete description of the noise characteristics, such as the noise power spectrum or the noise autocorrelation function, whereas others give only a partial description, such as rms noise. It will develop that the noise fluctuation in one pixel of a CT reconstruction is not independent of the noise fluctuations in other pixels. Rather, the fluctuations in two Separate pixels are, on the average, correlated. Random Noise: Image mottling, or fluctuations in the image density that change from one image to the next in an unpredictable and random manner, may be termed random noise. The Radiologist is familiar with random noise in the form of radiographic mottle found in standard radiographs taken with fast screen-film combinations [4]. Statistical Noise: The energy in x-radiation is transmitted in the form of individual chunks of energy called quanta. Hence the response of an x-ray detector is actually the result of detecting a finite number of x-ray quanta. The number of

Page 86

S. Sivakumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 86-91

detected quanta will vary from one measurement to the next, not because of inadequacies in the detection apparatus, but because of statistical fluctuations that naturally arise in the "counting" process. As more quanta are detected in each measurement, the relative accuracy of each measurement improves. Statistical noise in x-ray images arises from the fluctuations inherent in the detection of a finite number of x-ray quanta. Statistical noise may also be called quantum noise and is often referred to as quantum mottle in film radiography. Statistical noise clearly represents a fundamental limitation in x-ray radiographic processes. The only way to reduce the effects of statistical noise is to increase the number of detected x-ray quanta. Normally this is achieved by increasing the number of transmitted x rays through an increase in dose. Electronic Noise: In processing electric signals, electronic circuits inevitably add some noise to the signals. Analog circuits, those which process continuously varying signals, are most susceptible to additional noise. The difficulty of noise suppression is compounded by the fact that for some types of x-ray detectors, the electronic signals are very small. Digital circuits, those which process discrete signals as in digital computers, are relatively impervious to electronic noise problems[5]. Structural Noise: Density variations in the object being imaged that interfere with the diagnosis are sometimes referred to as structural "noise" or structural clutter. In standard radiography a large amount of structural clutter is produced by the superposition of various anatomic structures, for example, the image of rib bones overlaps that of the lung in a standard chest radiograph. The CT technique eliminates most of this superposition, but the radiologist should be aware that partial contributions may be introduced by structures that principally appear in adjacent CT slices. Some organs, such as the liver, may have density variations within them that have the appearance of' random noise. Although the texture pattern of the organ may not be reproducible from one CT scan to the next because of patient motion, this type of structural variation is, of course, not random. Indeed, the classification ofâ&#x20AC;&#x2122; this density variation as a type of noise is ill-advised, since the variation is intrinsic to the object itself. The study of the tissue texture may be interesting for its potential diagnostic value [6]. Round-Off Errors: Although digital computers are not subject to electronic noise, they do introduce noise in the reconstruction process through round-off errors. The errors arise from the limited number of bits used to represent numbers in the computer. For example, the product of two numbers must be rounded off to the least significant bit used in the computerâ&#x20AC;&#x2122;s representation of the number. Round-off errors can normally be kept at an insignificant level either through choice of a computer with enough bits per word or through proper programming. It should be pointed out that in some CT scanners the final reconstruction is stored with the least significant bit equal to one CT number (0.1 % of the linear attenuation coefficient of water). This should not influence the accuracy significantly so long as the rms noise is greater than one CT number [7]. III.

Denoising

Denoising plays a very important role in the field of the medical image pre-processing. It is often done before the image data is to be analyzed. Denoising is mainly used to remove the noise that is present and retains the significant information, regardless of the frequency contents of the signal. It is entirely different content and retains low frequency content. De-noising has to be performed to recover the useful information. In this process much attention is kept on, how well the edges are preserved and how much of the noise granularity has been removed [4-5] the main purpose of an image denoising algorithm is to eliminate the unwanted noise level while preserving the important features of an image. Noise Removal based on filtering models: Many image processing algorithms cannot work well in noisy environments. Specifically for the removal of noise from an input image there are several filters that can be considered as the state-of-art methods given their impressive performance [13]. There are many filters that are used to remove impulse noise from digital corrupted images. The following section describes the various filtering approaches considered in this work. Mean filter: This is the simplest of the mean filters. Let Sxy represent the set of coordinates in a rectangular subimage window of size m X n, centered at point(x, y). The arithmetic mean filter filtering process computes the average value of the corrupted image g(x, y) in the area defined by Sxy. The value of the restored image f at any point (x, y) is simply the arithmetic mean computed using the pixels in the region defined by Sxy. In other words,

This operation can be implemented using a convolution mask in which all coefficients have value 1/mn. An example of mean filtering of a single 3x3 window of values is shown below.

Page 87

S. Sivakumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 86-91

5+3+6+2+1+9+8+4+7=45/9=5 *

Mean filter simply smooth local variations in an image. Noise is reduced as a result of blurring. The mean filter is a simple sliding-window spatial filter that replaces the center value in the window with the average (mean) of all the pixel values in the window. The window, or kernel, is usually square but can be any shape[8][9]. Median filter: The best known order-statistics filter is the median filter, which, as its name implies, replaces the value of a pixel by the median of the gray levels in the neighborhood of that pixel:

The original value of the pixel is included in the computation of the median. An example of median filtering of a single 3x3 window of values is shown below. 6

0, 2, 3, 3, 4, 6, 10, 15, 97 * * * *

Center value (previously 97) is replaced by the median of all nine values (4). Median filters are quite popular because, for certain types of random noise, they provide excellent noise-reduction capabilities, with considerably less blurring than linear smoothing filters of similar size. Median filters are particularly effective in the presence of both bipolar and unipolar impulse noise. The median filter is also a sliding-window spatial filter, but it replaces the center value in the window with the median of all the pixel values in the window. As for the mean filter, the kernel is usually square but can be any shape. Midpoint filter: The Midpoint filter blurs the image by replacing each pixel with the average of the highest pixel and the lowest pixel (with respect to intensity) within the specified window size. For example, given the grayscale 3x3 pixel neighborhood; 22

150

158

219

The center pixel would be changed from 77 to 109 as it is the midpoint between the brightest pixel 219 and the darkest pixel 0 within the current window. The midpoint filter simply computes the midpoint between the maximum and minimum values in the area encompassed by the filter:

This filter combines order statistics and averaging. This filter work best for randomly distributed noise, like Gaussian or uniform noise.

Page 88

S. Sivakumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 86-91

Wiener Filter: The Wiener filtering executes an optimal tradeoff between inverse filtering and noise smoothing. It removes the additive noise and inverts the blurring simultaneously. Wiener filter estimates the local mean and variance around each pixel [8][9][10].

and

Where is the N-by-M local neighborhood of each pixel in the image, then creates a pixel-wise wiener filter using these estimates,

Where

the noise variance is not given, then the average of all the local estimated variances.

IV. Proposed Filtering Methods L-1(Leave-One) mean filter: This is the modified version of the mean filters. Let Sxy represent the set of coordinates in a rectangular subimage window of size m X n, centered at point(x, y). This filter computes the average value of the corrupted image g(x, y) in the area defined by Sxy with the interval of 1. The value of the restored image f at any point (x, y) is simply the arithmetic mean computed using the pixels in the region defined by Sxy. An example of mean filtering of a single 3x3 window of values is shown below. 5

5+6+4+8+7=30/5=6 *

L-1(Leave-One) median filter: This is the modified version of the median filters. Let Sxy represent the set of coordinates in a rectangular subimage window of size m X n, centered at point(x, y). This filter computes the median value of the corrupted image g(x, y) in the area defined by Sxy with the interval of 1. The value of the restored image f at any point (x, y) is simply the arithmetic mean computed using the pixels in the region defined by S xy. An example of L-1 median filtering of a single 3x3 window of values is shown below. 5

4,5,6,7,8 *

Page 89

S. Sivakumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 86-91

L-1(Leave-One) midpoint filter: This is the modified version of the midpoint filters by replacing each pixel with the average of the highest pixel and the lowest pixel (with respect to intensity) within the specified window size with L1 property. For example, given the grayscale 3x3 pixel neighborhood; 5

Low=4,Max=8,Midpoint=6 *

Evaluation metrics: The Mean Square Error (MSE) and the Peak Signal to Noise Ratio (PSNR) are the two error metrics used to compare image reconstruction quality. The higher value of PSNR and SNR denotes the better the quality of the reconstructed image [11]. The MSE represents the cumulative squared error between the reconstructed and the original image, whereas PSNR represents a measure of the peak error.

The lower the value of MSE and RMSE indicates the lower of the error. V. Experimental Results and Analysis Dataset: The Lung Image Database Consortium image collection (LIDC-IDRI) consists of diagnostic and lung cancer screening thoracic CT scans with marked-up annotated lesions. It is a web-accessible international resource for development, training, and evaluation of computer-assisted diagnostic (CAD) methods for lung cancer detection and diagnosis. The LIDC-IDRI collection contained on The Cancer Imaging Archive (TCIA) is the complete dataset of all 1,010 patients which includes all 399 pilot CT cases plus the additional 611 patient CTs and all 290 corresponding chest x-rays. The lungs image data, nodule size list and annotated XML file documentations can be downloaded from the National Cancer Institute website [12]. For the experiment we taken 40 Non-Cancer Lung CT scan images and 50 Cancer Lung CT images from the LIDC dataset. Figure 1: Different Filter Results on a Lung Cancer CT scan image

(a) Original

(e) Wiener applied

(b) Mean Filter applied

(f) L-1 Mean applied

(d) Midpoint applied

(g) L-1 Median applied (h) L-1 Midpoint applied

Page 90

S. Sivakumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 86-91

Table I: Comparison of different filters against the Lung CT scan images Image Filter Mean Median Midpoint Wiener L-1 Mean L-1 Median L-1 Midpoint

SNR 0.0410 0.0084 0.0680 0.0128 0.0449 0.0081 0.0647

Non-Cancer Image (average value of 40 images) PSNR MSE 76.2123 0.0016 79.4146 0.0009 72.5220 0.0037 83.2099 0.0004 75.5737 0.0019 78.8604 0.0010 72.4172 0.0038

RMSE 0.0399 0.0286 0.0607 0.0183 0.0429 0.0305 0.0615

Cancer affected Image (average value of 50 images) SNR PSNR MSE 0.0611 75.6819 0.0029 0.0223 79.0497 0.0023 0.0850 72.0057 0.0055 0.0242 82.6539 0.0007 0.0636 75.0489 0.0032 0.0198 78.4121 0.0023 0.0822 71.9250 0.0054

RMSE 0.0464 0.0348 0.0680 0.0217 0.0492 0.0365 0.0682

From the Table-I, Wiener filter approach gives better result compare with other filters in the values of high PSNR and low RMSE value for both the cancer and non-cancer images. For the cancer and Non-cancer images both the basic midpoint filter and the proposed Leave-One (L-1) midpoint filter performs at the same level. In performance wise, the wiener filter gives best noise elimination compare with others. Figure 2: Mean Absolute Error comparison of different filters

From the figure 2, the Mean Absolute Error also shows that the wiener filter performs superior than others. The proposed L-1Median modified filter performs more or less equal to wiener in both cases of the CT scan images. VI. Conclusion Image denoising is an essential task in medical image processing. Filters are very useful for removing noise from the images. In this paper we describe different filtering techniques and three more proposed filters. For the Lung CT scan images Wiener filter performs very well compare with other filters. The proposed L-1 filters are performing the same level of noise removal on both of the Cancer and Non-Cancer images compare with the basic filters, with less computational time. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

Shepp, L. A., and Logan, B. F.: The Fourier reconstruction of a head section, IEEE Trans. Nucl. Sci. NS-21:21-43, 1974. Huesman, R. H.: Analysis of statistical errors for transverse section reconstruction, Lawrence Berkeley Laboratory report #4278, 1975, University of California, Berkeley, Calif. Hanson, K. M., and Boyd, D. P.: The characteristics of computed tomographic reconstruction noise and their effect on detectability, IEEE Trans. Nucl. Sci. NS-25:160-173, 1978. Ter-Pogossian, M. M.: The physical aspects of diagnostic radiology, New York, 1967, Harper & Row, Publishers. Cohen, G.: : Contrast-detail-dose analysis of six different computed tomographic scanners, J. Comput. Assist. Tomogr. 3:197203, 1979. Pullan, B. R., Fawcitt, R. A., and Isherwood, I.: Tissue characterization by an analysis of the distribution of attenuation values in computed tomography scans: a preliminary report, J. Comput. Assist. Tomogr. 2:49-54, 1978. Burgess, A. E., Humphrey, K., and Wagner, R. F.: Detection of bars and discs in quantum noise, Proc. SPIE Appl. Opt. Instr. in Medicine VII 173:34-40, 1979. R.C.Gonzalez and R.E. Woods, Digital Image Processing. 2nd ed. Englewood Cliffs, NJ: Prentice-Hall; 2002. A.K. Jain, Fundamentals of Digital Image Processing. Englewood Cliffs, NJ: Prentice-Hall; 1989 Jingdong Chen,Jacob Benesty,Yiteng Huang, Simon Doclo,New Insights Into the Noise Reduction Wiener Filter, IEEE Transactions on audio, speech, and language processing, vol. 14, no. 4, July 2006. Z. Wang, A.C. Bovik, “A universal image quality index”, IEEE Signal Processing Letters, vol. 9, no. 3, pp.81-84, 2002. https://wiki.cancerimagingarchive.net/display/Public/LIDC-IDRI Sivakumar, S and Chandrasekar C, Lung Nodule Detection using Fuzzy Clustering and Support Vector Machines, IJET, vol. 5, no. 1, pp.179-185, 2013.

VII.

Acknowledgments

The First Author extends his gratitude to UGC as this research work was supported by Basic Scientist Research (BSR) Non-SAP Scheme, under grant reference number, F-41/2006(BSR)/11-142/2010(BSR) UGC XI Plan. The heading of the Acknowledgment section and the References section must not be numbered.

Page 91

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

Figure 1.

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net

A Novel Bit Error Rate Reduction Method for 3GPP-LTE-SCFDMA Using the Multiwavelet Transform Raad Farhood Chisab1,2 , Prof. (Dr.) C. K. Shukla3 1 Foundation of Technical Education, IRAQ 2 Dept. of ECE, SHIATS (Deemed to be University), Allahabad, UP, INDIA 3 Prof. at Dept. of ECE, SHIATS (Deemed to be University), Allahabad, UP, INDIA _________________________________________________________________________________________ Abstract: The rapid development of wireless communication technology has brought great convenience to people's lives and work. The goal of next generation of mobile wireless communication system is to achieve ubiquitous, high quality, high-speed mobile multimedia transmission with minimum bit error rate BER. To achieve this goal, the third generation partnership project 3GPP has developed the Long Term Evolution LTE. So, LTE has been considered as one of the core technologies of 4 th generation 4G wireless communication system. The LTE uses the Single Carrier Frequency Division Multiple Access SC-FDMA for uplink because of its ability to reduce BER and peak to average power ratio PAPR as compare to Orthogonal Frequency Division Multiple Access OFDMA that is used in downlink. In this paper a novel approach for implementing the SCFDMA was proposed using the Multiwavelet Transform MWT instead of Fast Fourier Transform FFT in order to reduce the BER and more robustness against fading channels. The new system was tested under six types of different channels cases. The results show that the proposed system gives lower BER as compare with the old system based on FFT. Keywords: 4G, LTE, SC-FDMA, BER, MWT, FFT __________________________________________________________________________________________ I. INTRODUCTION Nowadays, mobile radio system is immersed by more and more services with data rate from few Kbit/s up to several Mbit/s by an explosive growing demand for a wide variety of high quality of services in voice, video, and data. Wireless communications is moving rapidly towards small, low cost devices. However, the mobility and value of these devices is often limited by battery life since device miniaturization is progressing at a faster rate than battery technology optimization. Thus, the issue of battery life represents a key concern in the next generation of wireless communication systems [1]. In the next generation mobile communication systems, broadband data services are demanded. However, since the data rate increases, the transmit power should be increased to satisfy the required transmission quality, an unacceptably high transmit power is required. In addition to this, much higher transmit power is required in wireless communication systems due to the propagation path loss and the shadowing loss to guarantee the required quality of communication [2]. As the proliferation of smaller and faster devices increases, efficient use of limited battery resources becomes ever more paramount [3]. In the Orthogonal Frequency Division Multiplexing (OFDM), a lot of modulated sub-carriers are multiplexed in time domain, which causes high Peak-to-Average Power Ratio (PAPR). High PAPR may be major drawback of OFDM. Especially, it is more problematic in uplink than downlink because it incurs expensive mobile terminal or reduced uplink coverage [4]. Third generation partnership project long term evolution (3GPP LTE) represents a key advance in cellular mobile technology. The overall target of 3GPP LTE is to provide improved services, increase data rates, and higher spectral efficiency as well as lower latency [5]. SC-FDMA, which combines the features of the Single Carrier Frequency Domain Equalization (SC-FDE) and Frequency Division Multiple Access (FDMA) techniques, has been adopted as the standard for the uplink wireless access scheme in 3GPPLTE for its lower peak-to-average power ratio than that of orthogonal frequency division multiple access [6]. SC-FDMA can allocate different sub-carriers to different users to achieve multiple access interference free transmission. The simple single-tap equalizer can be used in the frequency domain for channel equalization, with Zero Forcing (ZF) or Minimum Mean Square Error (MMSE) criterion [5]. The scalable bandwidth of LTESC-FDMA is 1.5MHz to 20MHz while the number of subcarriers change from the 72 (for 1.5MHz) to 1200 (for 20MHz) [7]. II. SCFDMA As shown in figure 1, the transmitter of an SC-FDMA system converts a binary input signal to a sequence of modulated subcarriers. At the input to the transmitter, a baseband modulator transforms the binary input to a multilevel sequence of complex numbers xn in one of several possible modulation formats. The transmitter next groups the modulation symbols {xn} into blocks each containing N symbols. The first step in modulating the SC-FDMA subcarriers is to perform an N-point DFT to produce a frequency domain representation Xk of the input symbols. It then maps each of the N DFT outputs to one of the M (> N) orthogonal subcarriers that can be

Page 92

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

transmitted. If N = M/Q and all terminals transmit N symbols per block, the system can handle Q simultaneous transmissions without co-channel interference. Q is the bandwidth expansion factor of the symbol sequence. The result of the subcarrier mapping is the set (l = 0, 1, 2â&#x20AC;Ś, M-1) of complex subcarrier amplitudes, where N of the amplitudes are non-zero. As in OFDMA, an M-point IDFT transforms the subcarrier amplitudes to a complex time domain signal . Each then are transmitted sequentially [8].

Figure 1: The block diagram of the SCFDMA system There are M subcarriers, among which N (< M) subcarriers are occupied by the input data. In the time domain, the input data symbol has symbol duration of T seconds and the symbol duration is compressed to seconds after going through SC-FDMA modulation. There are two types of sub-carrier mapping which are Localized and Distributed mapping as shown in figure 2. In Localized Mapping the output from the DFT is mapped to a subset of consecutive subcarrier, confining only to a fraction of system bandwidth and the zero padding process is done either at the first or last, but the outputs of the DFT will be placed in the sequence order without any interchanging. In distributed Mapping the output of the DFT is assigned, non-continuously to the sub-carrier, over the entire bandwidth and the zero padding is done equally over the entire bandwidth [7]. The data block consists of N complex modulation symbols generated at a rate Rsource (symbols/sec). The N-point FFT produces N frequency-domain symbols that modulate N out of M orthogonal sub-carriers spread over a bandwidth W. The sub-carriers mapping process can be shown in figure 3. Where W can be defined as [9]: (1) Where F0 (Hz) is the sub-carriers frequency spacing. The channel transmission rate is: (Symbol/sec) (2) The bandwidth spreading factor Q is given by: (3) The SC-FDMA system can handle up to Q orthogonal source signals with each source occupying a different set of N orthogonal sub-carriers.

Figure 2: The two types of sub-carrier mapping For LFDMA, the frequency samples after subcarrier mapping

Figure 3: The process of sub-carriers mapping can be described as follows [8]:

Page 93

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

(4) Let (5) If q=0 then (6) If Then eqn. 5 can be expressed as follows: (7) (8) (9) As can be seen from above equations, LFDMA signal in the time domain has exact copies of input time symbols with a scaling factor of 1/Q in the N-multiple sample positions and in between values are sum of all the time input symbols in the input block with different complex-weighting. Now, For DFDMA, the frequency samples after subcarrier mapping can be described as follows. (10) Where Let Then (11) If q=0 then (12) (13) If

, since

Eqn. 19 can be expressed as follows after derivation (14)

From a resource allocation point of view, subcarrier mapping methods are further divided into static and Channel-Dependent Scheduling (CDS) methods. CDS assigns subcarriers to users according to the channel frequency response of each user. CDS is of great benefit with localized subcarrier mapping because it provides significant multi-user diversity which leads to improved system capacity and performance [6],[10]. This improvement can be shown in figure 6. For these reasons only LFDMA concept is proposed to use in the 3GPPLTE specifications. Therefore, we will focus and use this approach exclusively further in this paper. III. MULTIWAVELET TRANSFORM Multiwavelet Transform (MWT) was introduced previously and found wide spread application in several fields due to the orthogonally of basis functions and their greater suitability for use in communication systems. Multiwavelet are capable of reducing the Inter-Symbol Interference (ISI) and Inter-Carrier Interference (ICI), which are caused by the loss in orthogonality between the carriers [11]. Multiwavelet offer simultaneous orthogonality, symmetry, and short support that are not possible with the wavelet transform systems [12]. A very important Multiwavelet filter is the filter proposed by Geronimo, Hardian, and Massopust (GHM). For notational convenience, the set of scaling function and Multiwavelet function can be written by using the notation. , (15) The Multiwavelet studied to date are primary for r = 2. The GHM two scaling and wavelet function satisfy the following two scale dilation equation [13]. , Where

and

(16)

are matrix filter as

Page 94

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

, , Therefore the matrix filters

(17) , , can be written as:

and

(18)

(19) ,

(20)

III.1 Calculation of 1D-DMWT The one-dimensional discrete Multiwavelet Transform 1D-DMWT can be computed by the following steps [14]: 1. The input signal should be of length N, where N must be power of 2. For example 4, 8 16 and so on. 2. Construct the transformation matrix (W) with dimension of (2N×2N) using the GHM low and high pass filter matrices as in equation 42. 3. For GHM system the preprocessing step can be applied on the input signal by repeating the input stream multiplied by a constant value ( ). The DMWT can be get now by applying the matrix multiplication between (W) which is (2N×2N) and the preprocessing input signal which is (2N×1) H0 0

H1 0

H2 H0

H3 H1

 H2 G0 0

 H3 G1 0

 0 G2 G0

 0

 

0 0

   

 0 0 0

 H0 0 0

 H1 0 0

  G1 G 2 0 G0

 G3 G1

 0 G3 G1

0 H2  0 0 G2

0 H3  0 0 G3

 0

   G0



(21)

III.2 Calculation of 1D-IDMWT The one Dimensional Inverse Discrete Multiwavelet Transform (1D-IDMWT) should be used to reconstruct the original signal from the transformed signal. The reconstruction matrix can be getting by transpose or inverse the transformation matrix W. To compute a single level 1D Inverse Discrete Multiwavelet Transform using over-sampled scheme of post processing, the following steps should be done [15]: 1. Re-arrange the row pair of the (2N×1) vector such that the row pairs 1,2 and 3, 4 …N-1, N to become 1, 2 and 5, 6…2N-3, 2N-2 and the row pairs N+1, N+2 and N+3, N+4 … 2N-1, 2N to become 3, 4 and 7, 8 …2N-1, 2N of the resulting matrix. For example, the rows 1, 2, 3, 4, 5, 6, 7 and 8 will become 1, 2, 5, 6, 3, 4, 7 and 8. 2. Multiply the reconstructed matrix WT (transpose of transformation matrix W) with the rearrange resulting matrix (2N×1) that gets from previous step. 3. Taking just the odd rows (1, 3, 5… N-1) from the resulting matrix and neglect the even rows (2, 4, 6… N). the resulting vector (N×1) is the original reconstructed vector. The goal of preprocessing is to relate the given scalar input signal of length N to a sequence of length-2 vectors to start the analysis algorithm and reduce noise effects. In the one-dimensional signals, the computational method for DMWT and IDMWT by an oversampled scheme of preprocessing is convenient and influential and further performance gains were made by looking into alternative orthogonal bases functions and finding a better transform than FFT [16]. IV. PROPOSED MWT-SC-FDMA A block diagram of proposed LTE-SC-FDMA system is shown in figure 4 which illustrates the transmitter and receiver structure of MWT-LTE-SC-FDMA. At the input to the transmitter, a baseband modulator transforms the binary input to a multilevel sequence of complex numbers in one of several possible modulation formats. Commonly used baseband modulation schemes in upcoming LTE standard include QPSK, 16-QAM and 64-QAM. In general, transmitter adopts the modulation scheme to match the particular channel conditions and characteristics for the certain time instance. The process of modulation scheme is controlled by the resource allocation module. In this module the information is taken from the condition of the channel. If the channel in bad condition then the decision is to apply the low speed type of modulation such as QPSK, while if the channel in good condition then this module apply the high speed modulation types such as 16QAM or 64QAM. After choosing the suitable type of modulation, the transmitter concatenates the modulation symbols into blocks through Serial to parallel converter (S/P) block. After that the coded streams enter to new and important block which is the preprocessing block.

Page 95

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

The goal of preprocessing is to relate the given input signal of length N to a sequence of length-2 vectors to start the analysis algorithm and reduce noise effects.

Figure 4: Proposed MWT-LTE-SC-FDMA system In the one-dimensional signals, the computational method for DMWT and IDMWT by an oversampled scheme of preprocessing (repeated row) is convenient, influential and further performance gains were made by looking into alternative orthogonal bases functions and finding a better transform than Fast Fourier Transforms FFT.The signal after that is input to the DMWT block. The main motivation for using Multiwavelet is the superior spectral containment properties of Multiwavelet filters over Fourier filters. This high degree of suitability is related to the finite support and self-similarity of the basis functions. The replacement of the Fast Fourier transform by the Multiwavelet transforms leads to overcoming several limitations and improves performance efficiency. The data stream after that inputs to the sub carrier allocation block in which each user will get its sub carrier’s amount and type of sub-carrier mapping according to the information that feedback from the receiver side. In this feedback all the information about the channel will be feed in order to select the right type of mapping that suitable for specific type of channel in order to reduce the Bit Error Rate (BER). After that the data input to the blocks of post processing and IDMWT. In this block all the data will be returned to the time domain. Then the data will be changed to serial mode by the parallel to serial module (P/S) then to Digital to Analog module (DAC). After that this analog signal will be input to radio frequency module to travel through the channel. The signal suffers from more than one types of degradation which are AWGN and fading which are flat fading and selective fading.

Figure 5: The process of channel equalization All the process in the transmitter will be inverted in the receiver in order to get the original signal with minimum BER and distortion. The first step in the receiver is RF stage then analog to digital block A/D. After that, the signal enter to serial to parallel block (S/P) then to the part of system that try to equalize the effect of degradation which is known as channel equalization as shown in figure. 5. The received signal is equalized in the frequency domain using the FFT block. After the equalization block the equalized signal is then transformed back to the time domain using the IFFT by the following steps: Let E(m) where (m=0, 1, 2…NFFT -1) denote the equalizer coefficient for the mth sub carrier, the time domain equalized signal K(n) can be expressed as: (22) Where The equalizer coefficients E(m) are determined to minimize the mean square error between the equalized signal and the original signal. The equalizer coefficients are computed according to the types of the frequency domain equalization (FDE) in two methods as follows [9]: A. The zero forcing (ZF) Equalizer is (23) B. The Minimum Mean Square Error (MMSE) Equalizer is (24)

Page 96

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

Where * denotes the complex conjugate, H(m) is the transfer function of the channel and is average energy-per-bit to noise power spectral density. Equalization will be used to eliminate the effect of ISI. From figure 7 it can be noticed that the MMSE method is better than the ZF method and give lower BER compared with other method. Therefore, in all tests and simulations for channel models, the MMSE method will be use. The signal after output from the IFFT Block become in the time domain. Then, in order to complete processing, the signal input to the block of preprocessing then to the discrete Multiwavelet transform (DMWT) to make the system work in the Multiwavelet domain instead of the Fourier domain in order to get the better performance with minimum BER. Now all the subcarrier will be input to the subcarrier de-mapping which arrange the subcarrier for each user according to the information comes from the resource allocation module. After the rearrange of the subcarrier the new configuration will be input to the block of the post processing following by the block of IDMWT in order to return the subcarrier to time domain. After all this process the signal will be back to serial form then to the demodulation block that depends on the resource allocation module which takes the information from the channel to select suitable type of modulation. V. Results and Discussion The proposed system (3GPP-MWT-LTE-SC-FDMA) was simulated and run using MATLAB package version 7.12. The behavior of the proposed system was monitored while change the parameters that effect on the performance of the system. These parameters are listed in table I. The characteristics of wireless signal changes as it travels from the transmitter antenna to the receiver antenna. These characteristics depend upon the distance between the two antennas, the path or paths taken by the signal, and the environment around the path. The term channel refers to the medium between the transmitting antenna and the receiving antenna [17]. The profile of received signal can be obtained from that of the transmitted signal if we have a model of the medium between the two. This model of the medium is called Propagation channel model. Propagation channel models are essential tools for simulation and testing of wireless transmission systems. The literature is extensive on this topic, and many standards have recommended channel models for specific propagation environments [18]. Some of these channels will be study and apply on the proposed system and calculate the BER and comparing with other system based on FFT. These channels are: Table I: The parameters for simulation Parameters System bandwidth Modulation types

Value 5 MHz QPSK

Carrier Frequency

2025 MHz

Sub-carriers spacing

15 KHz

Sub-carriers mapping

Localized

No. of DMWT points Channel equalization Bit rate Target BER Channel estimation

64 MMSE 5 Mbps 10-4 Perfect

Channel Types

SUI, COST 259, Cost 207, 3GPP TDL, ITU,WCDMA

V.1 SUI channel models Stanford University Interim (SUI) model is developed for IEEE 802.16 by Stanford University. It is used for frequencies above 1900 MHz in this propagation model, three different types of terrains or areas are considered. These are called as terrain A, B and C. Terrain A represents an area with highest path loss, it can be a very dense populated region while terrain B represents an area with moderate path loss, a suburban environment. Terrain C has the least path loss which describes a rural or flat area. It is obvious that there are many possible combinations of parameters to obtain such channel descriptions. A set of six typical channels (SUI1, SUI2â&#x20AC;ŚSUI6) was selected for the three terrain types [19]. The performance of the system can be shown in figure 8. V.2 COST 259 channel models The European Co-Operation in the field of Scientific and Technical research (COST 259) was developed by the European COST 259 project. The COST 259 is wideband and capable of providing channel impulse responses in both spatial and temporal domains. It can also provide these in vertical and horizontal polarization components. It operates at the frequency range from 0.45 to 5 GHz. One of the work items identified in COST 259 is to propose a new set of channel models which overcome the limitations in the GSM channel models, while aiming at the same general acceptance. The main difference between the COST 259 model and previous models is that it tries to describe the complex range of conditions found in the real world by distributions of channels rather than a few typical cases.

Page 97

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

There are three types of channel models which are the Rural Area channel model (RAx), the Hilly Terrain channel model (HTx) and the Typical Urban channel model (TUx) [20]. The performance of the system can be shown in figure 9. V.3 COST 207 channel models The COST 207 model gives normalized scattering functions, as well as amplitude statistics for four typical environments which are rural area (RA), typical urban area (TU), bad urban area (BU), and hilly terrain (HT). The COST 207 model was presented as an outdoor wireless channel model. This model specifies power gains and time delays for four typical environments [21]. The performance of the system can be shown in figure 10. V.4 3GPP-TDL channel models The LTE standard adopts models based on the ITU-R M.1225 [18] recommendation and the 3GPP TS 05.05 [19] specification for GSM, widely used in the context of third generation mobile systems. The 3GPP models are defined by Tapped-Delay Line (TDL) models, where each tap corresponds to a multipath signal characterized by a fixed delay, relative average power and Doppler spectrum. They were designed for a 5 MHz operating bandwidth, and an apparent periodicity appears in their frequency correlation properties for higher bandwidths. The three models are the Extended Pedestrian-A (EPA), Extended Vehicular-A (EVA) and Extended Typical Urban (ETU) channel models [22]. The performance of the system can be shown in figure 11. V.5 ITU channel models For the selection of the air interface of third-generation cellular systems, the International Telecommunications Union (ITU) developed set of models that is available only as a tapped-delay-line implementation. The ITU models have been widely used, because they were accepted by international standards organizations. It specifies three environments: indoor, pedestrian (including outdoor to indoor), and vehicular (with high BS antennas). For each of these environments, two channels are defined: channel A (low-delay-spread case) and channel B (highdelay-spread case) [23]. ITU pedestrian profile B used for a mobile speed of 3km/h while ITU vehicular profile A used for a mobile speed of 60km/h [24]. The performance of the system can be shown in figures 12, 13 and 14. V.6 3G-WCDMA channels 3GPP UMTS (the Universal Mobile Telecommunications System) is the third generation (3G) uses Wideband CDMA (WCDMA or W-CDMA) to carry the radio transmissions. The scope of 3GPP was to produce globally applicable Technical Specifications and Technical Reports for a 3rd Generation Mobile Telecommunications System [25]. Since it was originally formed, 3GPP has also taken over responsibility for the GSM standards as well as looking at future developments including LTE and LTE Advanced. It contains three types of channels which are Indoor, Pedestrian and Vehicular. The performance of the system can be shown in figure 15.

Figure 6: performance under two types of mapping

Figure 8: system performance under SUI channel models

Figure 7: performance under two types of equalization

Figure 9: performance under COST 259 channels

Page 98

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

Figure 10: performance under COST 207 channels

Figure 12: performance under ITU-indoor channels

Figure 14: performance under ITU- Veh. Channels

Figure 11: performance under 3GPP-TDL channels

Figure 13: performance under ITU- Ped. channels

Figure 15: performance under WCDMA channels

VI. CONCLUSION In this paper two types of subcarrier mapping was tested and found that the localized subcarrier mapping was better than the distributed mode. Also two types of quantization was tested and found that the MMSE equalization is better than the ZF method also a new method of implementing the SC-FDMA was proposed based on the Multiwavelet Transform (MWT) instead of the Fast Fourier Transform (FFT). The proposed system gives the lower BER and high performance than the other system based on FFT. In all the channel types, which are six cases, it can be noticed that the reduction obtained in BER was different for each case according to channel types but still better than system based on FFT in all cases. This is a reflection to the fact that the orthogonal bases of the MWT are much significant than the orthogonal bases used in FFT. Thus, with all these tests for the channels and parameters it is conclude that the proposed system based on MWT works out better than the other system based on the FFT. Finally, as there is no cyclic prefix block in sending and receiving block, that means the proposed system is more bandwidth efficient. Thus high data rate transmission is possible without extra bandwidth, that means the quality and quantity of the signal reach to user will be improved.

Page 99

Raad F. Chisab et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 92-100

REFERENCES [1] [2]

[3] [4] [5] [6] [7]

[8] [9] [10]

[11] [12] [13] [14] [15]

[16] [17] [18] [19] [20] [21] [22]

[23] [24] [25]

P. Marilynn, Wylie-Green, and Erik Perrins, “a novel CPM-SC-FDMA transmission scheme for power efficient communication”, IEEE Global Telecommunications Conference, IEEE GLOBECOM 2008, pp.1 – 6, 2008. Masayuki Nakada, Kazuki Takeda, and Fumiyuki Adachi, “Channel Capacity Of SC-FDMA Cooperative AF Relay Using Spectrum Division And Adaptive Subcarrier Allocation”, 2nd IEEE International Conference on Network Infrastructure and Digital Content IC-NIDC2010, pp. 579 – 583, 2010. J. Dan and Abdullah Shami, “Energy Efficient Resource Allocation in SC-FDMA Uplink with Synchronous HARQ Constraints”, IEEE International Conference on Communications ICC, pp. 1 – 5, 2011. Byung Jang Jeong and Hyun Kyu Chung, “Pilot Structures for the Uplink Single Carrier FDMA Transmission Systems”, IEEE Vehicular Technology Conference VTC2008, pp. 2552 - 2556, 2008. Md. Masud Rana, “Performance of Sub-carrier Mapping in Single Carrier FDMA Systems under Radio Mobile Channels,” 13th IEEE International Conference on Computer and Information Technology ICCIT2010, pp. 175 – 180, 2010. Peng LI, Yu ZHU, Zongxin WANG, and Naibo WANG, “Peak-to-Average Power Ratio of SC-FDMA Systems with Localized Subcarrier Mapping,” IEEE Global Mobile Congress GMC2010, pp. 1 – 6, 2010. Dhirendra Kumar Tripathi, S. Arulmozhi Nangai, and R. Muthaiah, “FPGA Implementation of Scalable Bandwidth Single Carrier Frequency Domain Multiple Access Transceivers for the Fourth Generation Wireless Communication”, Journal of Theoretical and Applied Information Technology JATIT, ISSN: 1992-8645, Vol. 28, No. 2, 2011. Hyung G. Myung, “Single Carrier Orthogonal Multiple Access Technique for Broadband Wireless Communications”, Ph.D. Dissertation, Polytechnic University, January 2007. M. A. Abd El-Hamed, M. I. Dessouky, F. Shawki, Mohammad K. Ibrahim, S. El-Rabaie, and F. E. Abd El-Samie, “WaveletBased SC-FDMA System”, 29th National Radio Science Conference (NRSC 2012), pp. 447 – 460, 2012. Weidong Wang, Yan Zhou, Yuan Sang, Xue Shen, Fan Li, and Yinghai Zhang, “A Ue-Interfering Area Based Inter Cell Interference Coordination Scheme in SCFDMA uplinks”, 2nd IEEE International Conference on Network Infrastructure and Digital Content IC-NIDC 2010, pp. 681 – 686, 2010. Salih Mohammed Salih, “BER Reduction for DMT-ADSL Based on DMWT”, Cankaya University Journal of Science and Engineering, ISSN 1309 – 6788, Vol. 8, No. 2, 2011. Vasily Strela, Peter Niels , Gilbert Strang, Pankaj Topiwala, and Christopher Heil, “The Application of Multiwavelet Filter banks to Image Processing”, IEEE Transactions on Image Processing, Vol. 8, No. 4, pp. 548 – 563, 1999. M. Baro and Jacek Ilow, “Multi-band wavelet based spectrum agile Communications for Cognitive radio secondary user Communications”, IEEE International Symposium on broadband multimedia systems and broadcasting, pp. 1–5, 2008. Fritz Keinert, Wavelets and Multiwavelets, CHAPMAN & HALL/CRC Press Company, ISBN 1-58488-304-9, 2004. Abbas Hasan Kattoush, Waleed Ameen Mahmoud, Atif Mashagbah, and Ahmed Ghodayyah, “Multiwavelet Computed RadonBased OFDM Transceiver Designed and Simulation under Different Channel Conditions”, Journal of Information and Computing Science, ISSN 1746-7659, Vol. 5, No. 2, pp. 133-145, 2010. Jo Yew Tham, Lixin Shen, Seng Luan Lee, and Hwee Huat Tan, “A General Approach for Analysis and Application of Discrete Multiwavelet Transforms”, IEEE Transactions On Signal Processing, Vol. 48, No. 2, pp. 457 – 464, 2000. F. Molisch, Jinyun Zhang, and Toshiyuki Kuze, “Considerations for channel modeling in IEEE 802.16m,” IEEE C802.16m07/002, 2007. V. Erceg, “Channel Models for Fixed Wireless Applications”, IEEE 802.16 Broadband Wireless Access Working Group, IEEE 802.16.3c-01/29r4, 2001. Noman Shabbir, Muhammad T. Sadiq, Hasnain Kashif, and Rizwan Ullah, “Comparison of Radio Propagation Models For Long Term Evolution (LTE) Network”, International Journal of Next-Generation Networks (IJNGN), Vol.3, No.3, 2011. 3GPP TR 25.943 V6.0.0 (2004-12), 3rd Generation Partnership Project; Technical Specification Group Radio Access Networks; Deployment aspects (Release 6), 2004. Yao Xiao, “Orthogonal Frequency Division Multiplexing Modulation And Inter-Carrier Interference Cancellation”, MSc. Thesis, Louisiana State University, May 2003. Peral Rosado, Lopez Salcedo, Gonzalo Seco , Francesca Zanier, and Massimo Crisci, “Evaluation of the LTE Positioning Capabilities under Typical Multipath Channels”, 6th IEEE Advanced Satellite Multimedia Systems Conference ASMS, pp. 139 – 146, 2012. ITU-R M.1225, International Telecommunication Union, Guidelines for evaluation of radio transmission technologies for IMT2000, 1997. ITU-R M.2135-1, International Telecommunication Union, Guidelines for evaluation of radio interface technologies for IMTAdvanced, 2008. W. H. Tranter, K. S. Shanmugan, T. S. Rappaport, and K. L. Kosbar, Principles of Communication Systems Simulation with Wireless Applications, 1st edition, Prentice Hall Professional Technical Reference PTR , ISBN 0-13-494790-8, 2004.

Page 100

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A Novel Method for Image Segmentation Using Fuzzy Threshold Selection 1

Janakiraman.S, 2J.Gowri Pondicherry University, Puducherry, India Research Scholar, Bharathiar University, Coimbatore, Tamil Nadu, India _________________________________________________________________________________________ Abstract: The fuzzy technique is an operator introduced in order to simulate at a mathematical level the compensatory behavior in process of decision making or subjective evaluation. The following paper introduces such operators on hand of computer vision application. The areas of this work are in electronics and telecommunication engineering, which are very wide fields. This work is intended to implement the edge detection for digital image, so that it may be carried out to a big contour (face) identification of an object (an image). In this paper a novel method based on fuzzy logic reasoning strategy is proposed for edge detection in digital images using 16 fuzzy edge templates that show the possible direction of the edge in the image and then calculating the divergence between the origins image and the 16 fuzzy templates. Calculation of the maximum of the divergence value between the 16 templates and the original image of the same size. Set a threshold and applying the morphological operators. After emerging the fuzzy logic concept, a lot of Researcher of image processing shifted their attention towards the fuzzy logic concept and its applicability in the field of image processing. Keywords: fuzzy logic, Edge detection, fuzzy sets, Threshold, Morphological operator. __________________________________________________________________________________________ I. Introduction Last few decades the volume of interest, research, and development of computer vision systems has increased enormously. Nowadays they appear to be present in almost every sphere of life, from surveillance systems in car parks, streets, and shopping centers, to sorting and quality control systems in the majority of food production. Thus, introducing automated visual inspection and measurement systems are necessary, especially for the two dimensional mechanical objects, [1:8]. Sankura and Sezginb list over 40 different thresholding techniques [9]. In the past few years, fuzzy logic emerged as a different yet powerful tool to decision making [10], [11], [12]. In 1965 Zadeh proposed the concept of fuzzy logic and soon fuzzy concept gained popularity in the image processing field. Many technique have been proposed by researcher for fuzzy logic based edge detection [13], [14], and [15]. Zhao [16], proposed an edge detection technique by dividing the image into 3-fuzzy partitions (regions) and then finding the maximum entropy to give the best edge. He also derived the necessary condition to maximize the entropy function. Based on these condition three-level thresholding is obtained. A very important role is played in image analysis by what are termed feature points, pixels that are identified as having a special property. Feature points include edge pixels as determined by the well-known classic edge detectors of PreWitt, Sobel, Marr, and Canny [17:21]. Recently there has been much revived interest [22, 23] in feature points determined by "corner" operators such as the Plessey, and interesting point operators such as that introduced by Moravec. [24, 25] Classical operators identify a pixel as a particular class of feature point by carrying out some series of operations within a window centered on the pixel under scrutiny. In the proposed work an algorithm based on fuzzy logic rules are developed for detecting edges from an image. In order to avoid the complexity, Selection of the 16 fuzzy templates, converting into the fuzzy domain from the original image, Increasing the border line of the image of existing rows and columns respectively, finding Hesitation degree or intuitionistic fuzzy index, Calculation of the maximum of the divergence value between the 16 templates and the original image of the same size let the original image denoted by this formed by taking the 3x3 matrix in the border matrix, Selecting the minimum divergence among the 16 divergence values, so we have to transforming back in the image pixel domain i.e in the interval [1 255] domain and set a threshold. II. Fuzzy logic technique A. Fuzzy Image Processing Fuzzy image processing is the collection of all approaches that understand, represent and process the images, their segments and features as fuzzy sets. The representation and processing depend on the selected fuzzy technique and on the problem to be solved. Fuzzy image processing has three main stages: image fuzzification, modification of membership values, and, if necessary, image defuzzification. The fuzzification and defuzzification steps are due to the fact that we do not possess fuzzy hardware. Therefore, the coding of image data (fuzzification) and decoding of the results (defuzzification) are steps that make possible to process images

Page 101

Janakiraman S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 101-105

with fuzzy techniques. The main power of fuzzy image processing is in the middle step (modification of membership values). After the image data are transformed from gray-level plane to the membership plane (fuzzification), appropriate fuzzy techniques modify the membership values. This can be a fuzzy clustering, a fuzzy rule-based approach, and a fuzzy integration approach and so on, B. Fuzzy Sets and Fuzzy Membership Functions The system implementation was carried out considering that the input image and the output image obtained after defuzzification are both 8-bit quantized; this way, their gray levels are always between 0 and 255. The fuzzy sets were created to represent each variable’s intensities; these sets were associated to the linguistic variables “Black”, Edge and “white”. The adopted membership functions for the fuzzy sets associated to the input and to the output were triangles, as shown in Fig. 1. Black

White

Figure 1 (a) Membership functions of the fuzzy sets associated to the input and to the output

Figure 1 (b) Membership functions of the fuzzy sets associated to the input and to the output The functions adopted to implement the “and” and “or” operations were the minimum and maximum functions, respectively. The Mamdani method was chosen as the defuzzification procedure, which means that the fuzzy sets obtained by applying each inference rule to the input data were joined through the add function; the output of the system was then computed as the loom of the resulting membership function. The values of the three memberships function of the output are designed to separate the values of the blacks, whites and edges of the image. III. Algorithm Description The general algorithm for image thresholding based on measure proposed above can be formulated as follows: (1) Reading the pixel of the image. (2) Number of row(n) and column(r) of an image is taken. (3) Selection of the 16 fuzzy templates. a=0.3; b=0.8; t1 = [a a a; 0 0 0; b b b]; etc., (distinct templates)

Page 102

Janakiraman S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 101-105

(4) (5) (6)

(7)

(8) (9)

Converting into the fuzzy domain from the original image. All value in the interval of [0 1] (ie maximum pixel/element of the image) Finding Hesitation degree or intuitionistic fuzzy index 0.2 (assume) Hpimage for each templates = c*(1-pimagefor each templates) Calculation of the maximum of the divergence value between the 16 templates and the original image of the same size let the original image denoted by this formed by taking the 3x3 matrix in the border matrix. Finite iteration by Selecting the minimum divergence among the 16 divergence values and is positioned at the center of the templates position for the edge image. Fuzzy domain image(i,j)=max n [min r Hpimage(i,j)] Edge image in the fuzzy domain matrix is transformed back in the image pixel domain i.e in the interval [1 255] domain (multiply by 255) Set a threshold, and applying the morphological operators of matlab. IV. Experiment

The proposed system was tested with different Images, its performance being compared the existing edge detection algorithms and it was observed that the outputs of this algorithm provide much more distinct marked edges and thus have better visual appearance than that of the standard existing one. Five 350 X 350 pictures are taken into consideration. The result of various pictures in Berkley dataset are shown below in Figure 2, 3 & 4. Figure 2. Rice.tif

Figure 3.

Figure 4

For various threshold value the pixels ranges are displayed in the figure 5. It can be observed that the output that has been generated by the fuzzy method has found out the edges of the image more distinctly as compared to the ones that have been found out by the â&#x20AC;&#x153;thresholdâ&#x20AC;? edge detection algorithm. Thus the Fuzzy rule based System

Page 103

Janakiraman S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 101-105

provides better edge detection and has an exhaustive set of fuzzy conditions which helps to extract the edges with a very high efficiency.

Figure 5. Threshold Value and Pixel Ranges V. Conclusion In this paper, we have proposed a very simple and small but a very efficient, fuzzy rule based edge detection algorithm which infuse the concepts of artificial intelligence and digital image processing. Comparisons were made without threshold edge detection method. Displayed results have shown the accuracy of the edge detection using the fuzzy rule based algorithm over the other threshold method. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

I.M. Elewa, H.H Soliman and A.A. Alshennawy. "Computer vision Methodology for measurement and Inspection: Metrology in Production area". Mansoura Eng. First conf. Faculty of Eng. Mansoura Univ., March 28-30, 1995, Pp. 473-444. A. A. Alshennawy, "Measurement and Inspection of Three Dimensional Objects Using Computer Vision System", Pd.D thesis, Mansoura University, Egypt, 2003. H. D. Hofmann, "Application of Intelligent Measurements with Metrical Image Processing for Quality Control", presented at the 5th International Conference, PEDAC' 92, Alexandria, EGYPT, December 1992. R.T. Chin., C.A. Harlow, "Automated Visual Inspection: A survey ", IEEE Trans. Pattern Anal. Machine Intell., Vol. PAMI-4, pp. 557-573, November 1982. P. Rummel, "GSS – A Fast, Model-Based Gray-Scale Sensor System for Workpiece Recognition", Proceed. of 8th International Conf. on Pattern Recognition, pp. 18-21, Paris, France, Oct. 27-31, 1986. B. G. Batchelor, D. A. Hill and D. C. Hodgson (Eds.), "Automated Visual Inspection", IFS Publications Ltd., UK, 1985. A. M. Darwish and A. K. Jain, "A Rule Based Approach for Visual Pattern Inspection ", IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 10, No. 1, pp. 56-68, January 1988. G. Maitre, H. Hugli, F. Tieche and J. P. Amann, “Range Image Segmentation Based on Function Approximation”, published at ISPRS90, Zurich, Sept. 1990. B. Sankur, M. Sezgin, “Survey over image thresholding techniques andquantitative performance evaluation,” Electron. Imaging vol.13, pp.146–165, Jun.2004. L. A. Zadeh, “Fuzzy sets,” Information and Control, 8: 1965, pp. 338-353. A. Kaufmann, “Introduction to the Theory of Fuzzy Subsets Fundamentals Theoretical Elements, Vol. 1. Academic Press, New York, 1975. L.C. Bezdek, “Pattern Recognition with fuzzy Objective Function Algorithm,” Plenum Press, New York, 1981. K. Cheung and W. Chan, "Fuzzy One –Mean Algorithm for Edge Detection," IEEE Inter. Conf. On Fuzzy Systems, 1995, pp. 2039- 2044. Y. Kuo, C. Lee, and C. Liu, "A New Fuzzy Edge Detection Method for image Enhancement," IEEE Inter. Conf. on Fuzzy Systems, 1997, pp. 1069-1074. S. El-Khamy, N. El-Yamany, and M. Lotfy, "A Modified Fuzzy Sobel Edge Detector," Seventeenth National Radio Science Conference (NRSC'2000), February 22-24, Minufia, Egypt, 2000. M. Zhao, A. M. N. Fu, and H. Yan, “A Technique of ThreeLevel Thresholding Based on Probability Partition a Fuzzy 3- Partition”. IEEE Trans. on Fuzzy Systems, vol.9, no.3, June 2001, pp. 469- 479. G.I. Sanchez-Ortiz, A. Noble,“Fuzzy clustering driven anisotropic diffusion: enhancement and segmentation of cardiac MR images”, Nuclear Science Symposium, Vol. 3 , 1998, pp. 1873 -1874.

Page 104

Janakiraman S. et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 101-105 [18] [19] [20] [21]

[22] [23] [24] [25]

R.L. Webber, U.E. Ruttimann, and R.A.J. Groenhuis, "Computer correction of projective distortions in Dental Radiographs", J. Den. Res., vol. 63, No. 8, pp.1032-1036, August 1984. Song Wang, Feng Ge, Tiecheng Liu, "Evaluating Edge Detection Through Boundary Detection", Department of Computer Science and Engineering, University of South Carolina, Columbia, SC 29208, USA, 2006. Ivan Christov, " Multiscale Image Edge Detection",1.130/18.327, Spring 2004, Final Project, May 12, 2004. Nor Ashidi Mat Isa, " Automated Edge Detection Technique for Pap Smear Images Using Moving K-Means Clustering and Modified SeedBased Region Growing Algorithm", International Journal of The Computer, the Internet and Management Vol. 13.No.3 (SeptemberDecember,2005) pp 45-59. C.C. Leung, F.H.Y. Chan, K.Y. Zee, P.C.K. Kwok, "Compensation of bending errors in Intra-oral Radiographs using Block-byBlock Image Scaling", IEEE Trans. On Biomedical Engineering. (In manuscript). C.C. Leung, P.C.K. Kwok, K.Y. Zee, F.H.Y. Chan, and S.T.F. Lo, "Minimizing the bending error in Intra-oral Radiographs using Point-byPoint Interpolation with image scaling", Proceedings of the EMBEC'99, Part II, pp.1050-1051, Nov. 1999. Vienna. W.F. Chen, X.Q. Lu, J.J. Chen, and G.X. Wu, "A new algorithm of edge detection for color image: Generalized fuzzy operator", Science in China (Series A), Vol.38, No.3, pp.468-473, 1998. C.C. Leung, F.H.Y. Chan, W.F. Chen, P.C.K. Kwok and K.Y. Lam, “Thyroid Cancer Cells Boundary Location by a Fuzzy Edge Detection Method” ICPR’2000, Sept. 2000.

Page 105

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A Hybrid Approach to normalize the light illumination in facial images using DCT and Gamma Transformations C.Arunkumar1, T.Raghuram2, M.N.Sekharan3 Assistant Professor (Sr.Gr)1, Department of Information Technology, Amrita School of Engineering, Coimbatore, INDIA. Abstract: In this paper we propose a hybrid approach to normalize the illumination intensity in facial images using Discrete Cosine Transform (DCT) and gamma transformation. The image is transformed into frequency domain using DCT and the proposed method is used to get the areas affected by illumination. Those areas are normalized using power law transformation, using the gamma value found by the method. The experiment is done on the Yale B database and the results show an improvement in extracting edges from faces due to illumination normalization. Keywords: DCT, gamma transformation, illumination separation, region based separation) I.

Introduction

Over the last two decades, the research in the field of image processing has undergone an exponential growth with the ability to solve real world problems. That may include tracking down a vehicle in an image, face tracking using facial recognition, satellite image processing, and authentication in applications such as debit/credit cards, passports, voter ID etc. Among them there are many applications which use the image of the user as the input and perform various operations on them. This includes face identification, emotion detection, age estimation etc. In all these applications, the input is a face image. Processing facial images taken in an outdoor environment with variation in expression and illuminations are difficult. In previous works, the images were represented using edge maps, filters (Gaussian, Sobel, canny etc.), first or second derivatives and logarithmic transforms. But none of them were able to overcome the problem with the variation in illumination on the subject. II.

Earlier Works

A lot of work has been done for image enhancement by Vimal et al. [1]. They introduced an approach for finding gamma value. The calculation is based on the histogram peak value, which gives the details about the background color. This method works well for images, where background and foreground peaks are not well separated, i.e. the input image is of low contrast. But for the images with high contrast (a wide separation of foreground and background peaks), the method is not so effective. Michelle M. Mendonca et al. [2] showed an effective method for face recognition. It includes Log Transformation, Homomorphic filtering and Wavelets. Then face recognition is done using Principal Component Analysis (PCA). Of these three methods, it is found out that the wavelet method is very effective because the face image resulting from wavelet process enhances the edges and provide other details that facilitate further face identification process. S Venkatesan et al. [3] investigated the same idea described by Michelle M Mendoca et al. They used genetic algorithm instead of PCA for the recognition process. Shan et al. [4] investigated several illumination normalization methods and proposed Gamma Intensity Correction (GIC), Region-based strategy combining GIC with Histogram Equalization and Quotient Illumination Relighting (QIR) method. It is found that QIR is a better solution and that can estimate the lighting modes of an image. Ching Chung Huang et al. [5] proposed an algorithm that uses Homomorphic filtering. In this method, the face will be divided vertically and horizontally into equal halves. Then, Homomorphic filtering is done on those sub images and merged once again (two horizontal halves together and two vertical halves together, then they are merged together). Then an illumination reference model is constructed from the input face image and it is used to adjust the illumination. The combination of modified illumination information and facial feature is used to get normalized face image. Virendra P. Vishwakarma et al. [6] showed the usage of DCT to find the illumination of the image using the DC and AC coefficients. The image is then rescaled to lower frequency DC coefficients to get the image normalized. The operation is done on the full image. This is a modified version of the work by W. Chen et al. [7]. III.

Algorithm Description

A. Histogram Equalization It is a process of enhancing the contrast of an image using its histogram. The histogram of the digital image of gray scale value (range 0 to (L-1)) is defined by the discrete function

Page 106

C.Arunkumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 106-112

(1) Basically p ( ) is the probability of occurrence of each gray level in the image. Through this technique, the intensity is distributed over the entire image. Histogram equalization is defined as the transformation of input intensity levels ( ) into output intensity levels ( ) as (2) for k = 1 to L Figure 1: Histogram Equalisation (Left Side : Original Image, Right Side : Histogram equalized image)

B. Adaptive Histogram Equalization For images, which contain local regions with low contrast bright and dark regions, histogram equalization will not work globally. So we can use adaptive histogram equalization (AHE) for those images. Adaptive Histogram Equalization works by considering only small regions and based on their cumulative distribution function values, performs contrast expansion on these regions. So we can use this method for the facial images which are dark, where we canâ&#x20AC;&#x2122;t see the features like edges clearly. The image in fig 2 shows the use of this method. We can observe that the first face is considerably darker and features like nose, mouth and eyes are not clear. But after adaptive histogram equalization, we can observe their face parts more easily. As we can see, AHE does not influence the variation in the illumination. We can still use this to rebuild the image after finding the regions affected by the illumination. We will use DCT and their coefficients to find the illuminated regions. Figure 1: Output of AHE (a-input image, b-AHE output)

C. Discrete Cosine Transformation It is a popular technique in imaging and video compression. It converts the signals from spatial domain to the frequency domain. It has the property that most significant features of an image are concentrated in the few coefficients of the DCT. So it forms the major for lossy image compression algorithm known as JPEG (Joint Photographic Experts Group). The forward 2D DCT of an M x N block image is given by

(3) The inverse transform is defined as

(4) where

Page 107

C.Arunkumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 106-112

x and y are spatial coordinates in the image block in the spatial domain and u and v are coordinates in the frequency domain. Fig. 3 shows the property of the DCT coefficients in MxN block with the zigzag pattern used by JPEG compression to process the DCT coefficients. Figure 2: Zigzag pattern of choosing DCT coefficients

The C (0,0) is the DC coefficient, located in the top left corner of the DCT matrix. As the cosine of 0 is 1, C(0,0) is reduced as (5) The remaining ((MxN)-1) are AC coefficients. The DCT is performed on the entire image after pre-processing the images with the methods like histogram equalization or adaptive histogram equalization. D. Power Law Transformation It is a type of Image Enhancement which provides better contrast and a more detailed image as compared to nonenhanced image. The transformation is given by the equation (6) Figure 3: Plot of 'r' vs 's' for different values of gamma

The power law transformation is used for enhancing images for different types of display devices. The gamma for different displays is different. The Fig. shows the relation between the input (along x-axis) and the output gray levels (along y-axis) for different values of gamma. For example, the gamma for CRT lies between 1.8 and 2.5, which means the images present in the CRT, is dark. The equation is

Page 108

C.Arunkumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 106-112

(7) In [1], Vimal et al. showed that maximum contrast stretching occurs by choosing the value of for which the transformation function has the maximum slope at r= . That is, if m is the slope of the transformation, then should produce output which is produced when r = . So, he found out the formula for as follows, (8) IV.

Experimental Results and Discussion

The Fig 5 shows the flow of our approach. Our approach involves two major changes from the available methods. First step is finding the regions affected by the illumination from the input image. Second step is to adjust the illumination of these regions by finding the appropriate gamma for power law transformation. A. Face Database The experiment is done on the Yale B database which contains the image of 10 subjects each seen under 9 poses and 64 illumination types from different angles. In our approach, the concept is concentrated on the cropped face of those subjects. Figure 4: Proposed Approach

Discrete cosine transformation

Input Image

Extraction of illumination affected areas

Change the pixels of the illumination affected areas

Adaptive histogram equalization Finding an appropriate gamma value

Power Law Transformation

B. Adaptive Histogram Equalization This is the first step in our process. The input image is subjected to adaptive histogram equalization. Fig 2 shows the output of the adaptive histogram equalization. C. Discrete Cosine Transformation The modified version [6] of the method, which is mentioned in [7], shows the promising measure to take care of the illumination in the face. But it is seen that output given by [6] result has less brightness, thus reducing the details of the face. Fig 6 shows that the above conclusion is true. In [6], they rescaled the DC coefficients of low frequency and divided them by 50. This is done because the illumination variations are directly related to the lowfrequency DC coefficients. To deal with the problem mentioned above, we will multiply the constant 5 with the DC coefficients. Then IDCT is performed with the modified coefficients. Now this IDCT will give the image which represents the regions with illumination variance. This can be seen in Fig 7. Figure 5: Output of the method proposed in [6] with the brightness of the image affected

Page 109

C.Arunkumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 106-112

Figure 6: AHE output and IDCT output after multiplying DCT coefficients with the constant 5

D. Power Law Transformation Using the formula proposed in [1] for some gamma values it is observed that there is too much brightness or too little brightness. So in our approach, we set a constraint for the gamma values. Thus the intensity of the image is preserved with the appropriate gamma values. The condition is as follows IF Gamma > 5 THEN Gamma = 5 ELSE IF Gamma < 1 THEN Gamma = 1.02 END IF These ranges are set based on the observation from Fig. 4. With gamma greater than 5, there will be little brightness and too much brightness for gamma value less than 1. E. Final Step From the previous steps we have got the regions of the image affected by poor or too much illumination and a good gamma value for correcting that illumination. With this gamma value, we perform power law transformation on the original image (image ‘O’) and let the output of this transform be image ‘A’. Let the output of IDCT of DCT coefficients multiplied by 5 be image ‘B’ as shown in Fig 7. Now the output ‘R’ can be obtained by the following procedure. The pixel positions are indicated by (i, j) where ‘i’ is the row and ‘j’ is the column. FOR ALL PIXELS IN THE IMAGE IF B(i,j) == WHITE THEN R(i,j) = O(i,j) ELSE R(i,j) = A(i,j) END IF END FOR Now the output is shown in Fig 8. We can see that, the brightness of the input image is preserved and the necessary regions are modified correctly. The intermediate outputs are shown in figure 9.

Page 110

C.Arunkumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 106-112

Figure 7: Output of proposed method

Figure 8: Output during intermediate steps

F. Discussion To test this approach, we used this as a pre-processing step in the project â&#x20AC;&#x201C; Age estimation using facial images. Fig. 10 and Fig. 11 shows that the result of applying enhanced canny edge detector on an illumination corrected image is better than the result obtained by directly applying the edge detector on the subject image. Enhanced canny edge detector is applied on these images to extract wrinkles. Figure 9: Comparison of performance of edge detectors before and applying using illumination normalization on the image of person (age 22). Image Source: FGnet Database

Figure 10 Comparison of performance of edge detectors before and applying using illumination normalization on the image of person (age 50). Image Source: FGnet Database

Page 111

C.Arunkumar et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 106-112

Conclusion

In this paper, we propose a hybrid approach to normalize the illumination in facial images using DCT and power law transformation. The modified DCT method is used to extract the regions which are affected by illumination. Then power law transformation with appropriate gamma value is used in image enhancement. Unlike many other methods, in our method, the regions which don’t need illumination correction will never be modified. VI. [1] [2] [3] [4] [5] [6]

[7]

References

S.P.Vimal and P.K.Thiruvikraman, “Automated image enhancement using power law transformations,” Sadhana, Vol.37, Part 6, December 2012, pp.739-745. M.Mendonca, G.Denipote, A.S.Fernandes and V.Paiva, “Illumination Normalization Methods for Face Recognition.” S.Venkatesan and S.Srininvasa Rao Mandane, “Experimental study on illumination normalization methods for effective face identification by genetic algorithm,” Journal of Global Research in Computer Science, vol.3, no.2, February 2012. S.Shan, W.Gao, B.Cao and Zhao, “Illumination Normalization for Robust Face Recognition Against Varying Lighting Conditions,” Proc. Of the IEEE International Workshop on Analysis and Modelling of Faces and Gestures, vol.17, 2003, pp.157-164. C.Hung and Y.Liu, “Novel illumination – normalization method on region information,” Proc. Of SPIE-IS&T Electronic Imaging, vol.5672. V.P. Vishwakarma, S. Pandey and M. N. Gupta, “A Novel Approach for Face Recognition using DCT Coefficients Re-scaling for Illumination Normalization,” in Proc. Of IEEE Int. Conf. on Advanced Computing & Communication (ADCON 2007), Dec. 2007, pp. 535-539. Weilong Chen, Meng Joo and Shiquian Wu, “Illumination Compensation and Normalization for Robust Face Recognition Using Discrete Cosine Transform in Logarithm Domain,” IEEE Transactions on Systems, Man, And Cybernetics – Part B: Cybernetics, vol.36, No. 2, pp 458-466, Apr. 2006.

Page 112

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Review on Scuderi Split Cycle Engine Ashwini S. Gaikwad 1, Rajendra M.Shinde 2 P.G. Student, Automobile Engineering Department, RIT, Islampur, Sangli, Maharashtra, India. 2 Professor, Automobile Engineering Department, RIT, Islampur, Sangli, Maharashtra, India.

Abstract: Internal combustion engines are the important prime movers on the earth. The study of I.C. engine is an active field of research for many automobile industries and from environmental point of view. I.C. engine has wide application such as in automobile industries, transportation in sea as well as in air and for industrial purpose. Many technologies have been invented to enhance performance and to reduce emissions from the IC engines, the Scuderi Split Cycle Engine is one of them. This paper review, a new IC engine developed by Scuderi Group called ‘Scuderi Split Cycle Engine’. This engine is more efficient than a conventional engine and also has less emission. Keywords: Scuderi Split cycle engine, Internal Combustion Engine, Miller cycle, Otto cycle. I. Introduction Scuderi Split-cycle engines divide the four strokes of conventional engine such as intake, compression, power, and exhaust into two separate but paired cylinders. The first cylinder called as compression cylinder, which is used for intake and compression of air. The compressed air is then transferred through a crossover passage from the compression cylinder into the expansion cylinder. In expansion cylinder fuel is injected and burned to undergo combustion and exhaust stroke. A Scuderi split-cycle engine is an air compressor on one side and a combustion chamber on the other. The Backus Water Motor Company of Newark, New Jersey produced an example of a split cycle engine as far back as 1891. Various scientists and engineers worked to explore the possibility of the split cycle engine. But, none has matched the results obtained by Late Carmelo J. Scuderi. The Scuderi Group, engineering and licensing company based in West Springfield, Massachusetts and founded by Carmelo Scuderi’s children, completed the prototype and unveiled it to the public on April 20, 2009. II. The Component of Scuderi Split Cycle Engine 1. Intake and Exhaust Valves The intake and exhaust valves are inwardly opening, i.e., towards the piston. Intake valve controls the amount of air coming in the compression cylinder and exhaust valve rejects the exhaust gases from the power cylinder to the surrounding. Figure 1: Component of Split Cycle Engine

2. Compression and Expansion Cylinder The split-cycle engine design splits the four strokes and the processes of suction and compression take place in the compression cylinder, and the processes of expansion and exhaust take place in the power cylinder. The compression cylinder consists of the a piston, inlet valve and the crossover compression valve whereas a power cylinder consists of a piston, exhaust valve, crossover expansion valve and the spark plug. The efficiency of an internal combustion engine is increased if the gas is expanded more during the expansion stroke than it is compressed during the compression stroke. In Split Cycle Engine the design of compression cylinder is different from that of the expansion cylinder. The displacement volume of compression cylinder is less than that of the displacement volume of the power cylinder. This makes the expansion stroke greater than that of the compression stroke. Due to this arrangement more work is obtained in power stroke and less work has to be given in compression cylinder which is not possible in conventional IC engine. 3. Compression Piston and Expansion Piston A compressor cylinder including a compression piston moves within the compression cylinder and it is connected to the crankshaft. Compression piston reciprocates through the suction stroke and the compression stroke during a

Page 113

Ashwini S. Gaikwad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 113-117

single rotation of the crankshaft. An expander cylinder includes an expansion piston which moves within an expansion cylinder and is connected to the crankshaft such that the expansion piston reciprocates through an expansion stroke and an exhaust stroke during a single rotation of the crankshaft. 4. Crossover Passage A crossover passage interconnecting the compression and expansion cylinders, and it includes a crossover compression (XovrC) valve and a crossover Expansion (XovrE) valve. It also consists of a fuel injector. The crossover passage transfers the compressed air from the compression cylinder to the expansion cylinder and maintains the pressure in between the two cylinders. It consists of a fuel injector which injects the fuel in the passage. 5. Crossover Valve Due to very high compression ratios an outwardly opening (opening outward away from the cylinder and piston) crossover compression (XovrC) valve at the crossover passage inlet is used to control flow from the compression cylinder into the crossover passage. Due to very high expansion ratios an outwardly opening crossover expansion (XovrE) valve at the outlet of the crossover passage controls flow from the crossover passage into the expansion cylinder. 6. Crank and the Crankshaft The pistons of both the compression and power cylinder are connected to the crank by crankshaft. When the crank rotates it drives the piston in the power as well as in the compression cylinder. Generally the expansion piston leads the compression piston by 20 degrees. 7. Fuel Injector The main function of fuel injector is to inject the fuel and mixes it with air and it form homogeneous mixture of air and fuel which is necessary for the proper combustion of charge. 8. Spark Plug It provides the spark which is required to ignite the fuel and start the combustion process. It gives energy to the charge and increases the temperature of fuel to its ignition temperature. III. Scuderi Split Cycle Engine The four strokes of the Otto cycle are “split” over the two cylinders such that the compression cylinder, it perform the intake and compression strokes, and the expansion cylinder, perform the expansion and exhaust strokes. The working of the engine is explained in four strokes: Figure 2: Scuderi Split Cycle Engine

1. Suction Stroke During the intake stroke, intake air is drawn into the compression cylinder through an intake manifold disposed in the cylinder head. An inwardly opening intake valve controls fluid communication between the intake manifold and the compression cylinder. The intake air is supercharged at 1.5 bar. A boosting device is connected to the intake manifold such as supercharger. 2. Compression Stroke During the compression stroke, the compression piston pressurizes the air and, upon XovrC opening, drives the air to crossover passage. This means that the compression cylinder and compression piston are a source of high pressure gas to the crossover passage, which acts as the intake passage for the expansion cylinder. Fuel injector injects fuel into the pressurized air at the exit end of the crossover passage in correspondence with the XovrE valve opening, which occurs shortly before expansion piston reaches its top dead centre position. At this time, the pressure in the crossover passage is more than the pressure in the expansion cylinder. Fuel is injected in the Cross over valve. 3. Expansion Stroke When XovrE valve opens, the pressure in crossover passage is substantially higher than the pressure in expansion cylinder. This high pressure ratio causes initial flow of the air or fuel charge to flow into expansion cylinder at high speeds. As piston begins its descent from its top dead centre position, the XovrC valve is still open, and the spark plug, which includes a spark plug tip which is fired to initiate combustion in that region. The high speed flow of the air/fuel charge is particularly advantageous to split-cycle engine because it causes a rapid combustion

Page 114

Ashwini S. Gaikwad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 113-117

event, which enables the split- cycle engine to maintain high combustion pressures even though ignition is initiated while the expansion piston is descending from its top dead centre position. The XovrE valve is closed after combustion is initiated but before the resulting combustion event can enter the crossover passage. The combustion event drives the expansion piston downward in the power stroke. 4. Exhaust Stroke During the exhaust stroke, exhaust gases are pumped out of the expansion cylinder through exhaust valve disposed in cylinder head. An inwardly opening poppet exhaust valve, controls fluid communication between the expansion cylinder and the exhaust valve. IV. The Distinctive Element After Top Dead Centre Firing (ATDC Firing) Unlike conventional engine which are fired before top dead centre split cycle engine are fired after top dead centre, i.e., spark is generated when the piston starts its downward travel. To fire BTDC in a split-cycle engine, the compressed air is expanded into the power cylinder as the power piston is in upward stroke. The pressure of the compressed air is released and the work done takes place. To fire BTDC the power piston recompresses the air. Therefore engine required to perform the work of compression twice. Although considered bad practice in conventional engine design, firing after top dead centre in a split-cycle arrangement eliminates the losses created by recompressing the gas. When the engine is fired before top dead centre position, maximum combustion pressure is obtained at TDC. However at this position the connecting rod and the crankshaft throw are nearly aligned with the cylinder axis. Thus the pressure generated acts vertically downwards in line with the connecting rod. Whereas if engine is fired after TDC adjustments can be made to coincide maximum combustion pressure with maximum torque generated thus increasing the thermal efficiency. V. Air Standard Cycle The conventional SI engine works on the Otto cycle in the same way there is an air standard cycle for the Scuderi d split cycle engine which is Miller Cycle. A. Miller Cycle It is an over-expanded cycle, i.e., a cycle which has expansion ratio higher than its compression ratio. Apart from this it also fulfils the Miller effect. The efficiency of an internal combustion engine is increased if the gas is expanded more during the expansion stroke than it is compressed during the compression stroke. In the Miller cycle, this is typically accomplished by early inlet valve closing and late inlet valve closing. For example, if the inlet valve of a conventional engine is closed late a portion of the intake air that was drawn into the cylinder during the intake stroke is pushed back out of the cylinder from the intake port. The intake valve is kept open during about the first 20% to 30% of the compression stroke so the actual compression occurs only in about the last 80% to 90% of the compression stroke. A1. Miller Cycle Utilizing Early Inlet Valve Closing Referring to Fig No 3, the same effect can be achieved in the Miller cycle by early inlet valve closing. In such case, the pressure remains constant during the intake stroke from point 6 to point 1. Then at point 1 the intake valve closes, and the cylinder pressure decreases from point 1 to point 7. During the compression stroke, from point 7 to point 1 the pressure increases, replaces the previously traced path, and continues to point 2 during the remainder of the compression stroke. The net result of this is the same as late intake valve closing which is less than the entire piston stroke is effectively used for compression, decreasing the effective compression ratio. Figure 3: PV Curve for Early Inlet Valve Closing

Page 115

Ashwini S. Gaikwad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 113-117

A2. Miller Cycle Utilizing Late Inlet Valve Closing Figure 4: PV Curve for Late Inlet Valve Closing in Miller Cycle

From point 6-5-1 and 5-1, the intake stroke of the piston from top dead centre to bottom dead centre, the cylinder pressure follows a constant pressure line from point 6 through point 1 and finally to point 5. During the initial portion of the compression stroke, the cylinder pressure retraces the pressure line from point 5 back to point 1. At point 1 the intake valve closes and the cylinder pressure increases from point 1 to point 2 during the remainder of the compression stroke. The volume swept along the path 1-5 is cancelled by the path 5-1, and the effective compression ratio is the volume at point 1 divided by the volume at point 2 but for Otto cycle it is the volume at point 5 divided by the volume at point 2. From point 1-2, The compression piston pressurizes the air and, upon XovrC opening, drives the air into the crossover passage. The XovrC valve opens in between the compression stroke. Fuel injector injects fuel into the pressurized air at the exit end of the crossover passage in correspondence with the XovrE valve opening, which occurs shortly before expansion piston reaches its top dead centre position. At this time the pressure in the crossover passage is more than that in the expansion cylinder. From point 2-3, The charge usually enters the expansion cylinder shortly after expansion piston reaches its Top Dead Centre position (TDC), although it may begin entering slightly before TDC. As piston begins its descent from its top dead centre position, and while the XovrE valve is remain open, spark plug, which includes a spark plug tip that protrudes into cylinder is fired to initiate combustion in the region around the spark plug tip. From point 3-4,The XovrE valve is closed after combustion is initiated but in such a way that products of combustion doesn’t enter the crossover passage. The combustion event drives the expansion piston downward in the power stroke. From point 4-5, during this process the exhaust valve opens and the gases are released from the power cylinder when piston moves from its BDC to TDC. VI. Conclusion It can be seen that this engine is more efficient than conventional engine due to its inherent features, from the literature review. The Scuderi Split Cycle Engine development is a result of the endeavour to increase the efficiency and reduce the emissions of the engine. The research in the development of this new engine is very unique attempt as compared to the researches going around world for two decades for improving the power output by installing additional accessories to the conventional engine. The research and development of Scuderi Split Cycle Engine is very essential because of high requirement of efficient engines with very less emissions are prevalent. The performance of Scuderi split cycle engines are more efficient than a conventional engine. Due to high performance and efficiency, conventional engines are replaced by Scuderi split cycle engines for the same applications in power generation, automobiles. Therefore they are future alternatives of the conventional engines. VII. References [1]. [2]. [3]. [4]. [5]. [6]. [7]. [8].

Ganeshan,V.(2007) “Internal Combustion Engines, Third edition”, Tata McGraw-Hill, New-Delhi. John Heywood, (1988) “Internal Combustion Engine Fundamentals”, Tata McGraw-Hill. New-Delhi. Li. G & Sapsford, M, “CFD Simulation of DI Diesel Engine Combustion Using VECTIS”, Ricardo consulting engineer’s Ltd. UK.. Rajput R K (2007), Internal Combustion Engines, Laxmi Publications, New Delhi. Scuderi Group LLC,, Exhaust valve timing for split cycle engine, WO 2012/050910 A1 at www.ScuderiGroup.Com, 19 April 2012. Sudeer Gowd Patil1, Martin A.J, Ananthesha “Study on Performance Characteristics of Scuderi – Split Cycle Engine”, April 2012. Scuderi Group LLC (2011), Split -cycle Air-hybrid engine with air tank valve, EP Patent No. WO2011/115873 A1, at www.Scuderi Group.com, 22 September 2011. Scuderi Group LLC, “Split -cycle Air -hybrid engine with air expander and firing mode”,WO2011/115868A1,at www.ScuderiGroup.com, 22 September 2011.

Page 116

Ashwini S. Gaikwad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 7(1), December 2013February, 2014, pp. 113-117 [9]. [10]. [11]. [12]. [13]. [14]. [15]. [16].

Scuderi Group LLC, “Air supply for components of a split -cycle engine”, WO 2010/129872A1,at www.ScuderiGroup.com, 11 November 2010. Scuderi Group LLC, “Variable volume crossover passage for a split-cycle engine”, WO2010/120856A1,at www.ScuderiGroup.com, 21 October 2010. Scuderi Group LLC, “Part Load Control in Split Cycle Engine”, WO 2010/120499 A1, at www.ScuderiGroup.com, 21st October 2010. Scuderi Group LLC,, Crescent-Shaped Recess In Piston of a Split-Cycle Engine, WO2010/117713A1 at www.ScuderiGroup.Com, 14 October 2010. Scuderi Group LLC, “Scuderi- Engine Brochure”, at www.ScuderiEngine.com, 2010. Scuderi Group LLC (2009), Split Cycle Four Stroke Engine, EP Patent No. 1639247B9, available at www.Freepatentsonline.com, 14th January 2009. Scuderi Group LLC, “Split Cycle Engine With helical Crossover Passage”, WO 2009/020491 A1 at www.ScuderiGroup.com, 12th Feb 2009. Sapsford.S.M & Bardsley.E.A, “Mesh Generation: The Ricardo Philosophy”, Ricardo consulting engineers, UK.

Page 117