Instant download Health information science 4th international conference his 2015 melbourne australi
Health Information Science 4th
International Conference HIS 2015 Melbourne Australia May 28 30 2015 Proceedings 1st Edition Xiaoxia Yin
Visit to download the full and correct content document: https://textbookfull.com/product/health-information-science-4th-international-conferen ce-his-2015-melbourne-australia-may-28-30-2015-proceedings-1st-edition-xiaoxia-yin /
More products digital (pdf, epub, mobi) instant download maybe you interests ...
Information Theoretic Security 8th International Conference ICITS 2015 Lugano Switzerland May 2 5 2015 Proceedings 1st Edition Anja Lehmann
Codes Cryptology and Information Security First International Conference C2SI 2015 Rabat Morocco May 26 28 2015 Proceedings In Honor of Thierry Berger 1st Edition Said El Hajji
Trust Management IX 9th IFIP WG 11 11 International Conference IFIPTM 2015 Hamburg Germany May 26 28 2015 Proceedings 1st Edition Christian Damsgaard Jensen
LNCSSublibrary:SL3–InformationSystemsandApplications,incl.Internet/Web,andHCI SpringerChamHeidelbergNewYorkDordrechtLondon c SpringerInternationalPublishingSwitzerland2015
Peter Wlodarczak(), Jeffrey Soar, and Mustafa Ally
University of Southern Queensland, West Street, Toowoomba Qld 4350, Australia wlodarczak@gmail.com, {Jeffrey.Soar,Mustafa.Ally}@usq.edu.au
Abstract. There is increasing interest in Big Data analytics in health care. Behavioral health analytics is a care management technology that aims to improve the quality of care and reduce health care costs based capture and analysis of data on patient’s behavioral patterns. Big Data analytics of behavioral health data offers the potential of more precise and personalized treatment as well as monitor population-wide events such as epidemics.
Mobile phones are powerful social sensors that are usually physically close to users and leave digital traces of users’ behaviors and movement patterns. New Apps (application or piece of software) are emerging that passively collect and analyze mobile phone data of at-risk patients such as their location, calling and texting records and app usage, and can find deviations in a user’s daily patterns to detect that something is wrong before an event occurs. Data mining and machine learning techniques are adopted to analyze the “automated diaries” created by the smart phone and monitor the well-being of people. The App first learns a patients daily behavioral patterns using machine learning techniques. Once trained, the App detects deviations and alerts carers based on predictive models.
This paper describes the techniques used and algorithms for reality mining and predictive analysis used in eHealth Apps.
Keywords: Reality mining · Big data · Machine learning · eHealth · Predictive analytics · Behavioral health analytics · Mobile sensing
1 Introduction
An important question for behavioral epidemiology and public health is to better understand how individual behavior is affected by illness and stress [3]. Someone who becomes depressed isolates himself and has a hard time to get up and go to work. He shows deviations from his normal behavioral patterns. Smartphones produce significant amounts of behavioral data. They are essentially off-the-shelf wearable computers. They can provide a convenient tool for measuring social connectivity features related to phone calls and text messages [1]. Users usually keep mobile phones physically close to themselves. The mobile sensor data thus reflects the same movement patterns as the user. Most Smartphones are equipped with accelerometers for motion detection, GPS (Global positioning system) monitor where a user visits and call logs record call duration. They are powerful social sensors for spatio-temporal data. Decreased movement detected by motion sensors or infrequent texts in the message log might be symptoms of depression. Shorter than usual calls might signal isolation.
Real-time data collection and analysis of mobile phone data reveals information on the health state of a user and can be used to diagnose if a patient becomes symptomatic and prompt early treatment. Symptoms that can be detected are anxiety, stress, disease spread, and obesity [2, 3]. If symptoms are detected, a health care center can be alerted and a nurse can call the patient and check on his situation. This type of proactive healthcare is especially useful for high risk patients or patients susceptible of underreporting like mentally ill or elderly people.
2
Methodology
Reality mining refers to the process of collecting and analyzing machine sensed human behavioral data such as movement patterns, human interactions and human communication patterns, with the goal of detecting predictable behavioral patterns [18]. Reality mining comprises four phases. A data collection phase, a data preprocessing phase, a data mining phase and a post-processing phase. Sometimes a predictive analysis phase is added. Here the predictive step is considered part of the post-processing.
2.1 Data Collection
The data collection phase records a patient’s behavioral data from interactions from electronic exchanges (call records, SMS logs, email headers) and contextual data (location information). Sometimes other data like face-to-face proximity for individuals has been collected too using the mobile phones Bluetooth connection [2]. The mobile phone is used to extract conversational partners and location of a user, that is, the total number of interactions, the diversity of interactions, and the diversity (entropy) of his behavior [2]. Smartphones provide APIs to access the underlying functionality such as GPS sensors or call logs programmatically.
2.2 Data Pre-processing
Not all data collected is useful. The data has thus to be relevance filtered first. Also the raw sensor data is not in a format that can be used by most ML algorithms. The data has to be transformed into a feature vector. A feature in a feature vector can represent the coordinates of a location or the call duration. Eigenvector analysis, commonly known as principal components analysis, is the optimal linear method for obtaining a low-dimensional approximation to a signal such as observations of user behavior [5]. Behavioral structure can be represented by the principal components of the spatiotemporal data set, termed eigen-behaviors [10]. The term eigenbehavior was introduced by Eagle and Pentland [11]. We represent this behavioral structure by the principal components of the complete behavioral dataset, a set of characteristic vectors we have termed eigenbehaviors [11]. Eigenbehaviors provide an efficient data structure for learning and classifying tasks.
To calculate the Eigenbehavior a person’s behavior has to be measured, for instance the time sequence of their phone calls or text messages. For a group of M people, and the behaviors Γ1, Γ2, …, ΓM, the average behavior is:
A set of M vectors, Φi = Γi – Ψ, is defined to be the deviation of the normal behavior. Principle components analysis is subsequently performed on these vectors generating a set M orthonormal vectors, un, which best describes the distribution of the set of behavior data when linearly combined with their respective scalar values, λn [5]. The Eigenvector and Eigenvalues of the covariance matrix of Φ are calculated as:
Where the Matrix A = [Φ1, Φ2, …, ΦM]. A typical daily pattern is leaving the sleeping place in the morning, spending time in a small set of locations during office hours, and occasionally moving to a few locations in the evening and on the weekends. For typical individuals the top three Eigenbehavior components account for up to 96% of the variance in their behavior [5]. This means that a person’s location context can be classified with high accuracy.
2.3 Data Mining
To make predictions about a person’s health state, the behavioral data needs to be automatically classified into normal and deviant behavior. Machine learning (ML) techniques have been successfully applied for data classification problems. A given data set is typically divided into two parts: training and testing data sets with known class labels [8]. The class label is “normal” and “deviant” behaviour. The training data is the data collected during a training phase to learn a patient’s normal behaviour as represented by the Eigenvector. It is used to train a model. The test data, the realtime behavioural data, is then applied against the trained model. The model analyses the data for any abnormalities and makes predictions about the health state. Typical supervised learning methods include naïve Bayes classification, decision tree induction, k-nearest neighbors, and support vector machines [4]. There are many more ML algorithms. Experience shows that no single machine learning scheme is appropriate to all data mining problems [9]. Usually several algorithms are trained and compared to determine which one gives the most accurate results for a given problem [6]. Ultimately we want to obtain a decision function f, that classifies the behavioral pattern h as normal ( N), or deviant (D). If we denote the set of all behavioral patterns by H, we search for a function f:H → {N,D}. We use the set of behavioral data collected during the training phase {(h1, c1), (h2, c2), . . , (hn, cn)}, where: hi ∈ H, ci ∈ { N,
D}, to train the model. The naïve Bayes classifier is a family of simple probabilistic classifiers based on the Bayes theorem [6]. Decision tree learning creates decision
trees, where a decision could be: did the patient go to coordinate x,y early in the morning, yes/no. Support Vector Machine (SVM) classifications are based on statistical learning theory and classifies data by separating them with a hyperplane. Which classification algorithm performs best can depend on the type of illness, but other factors such as the patient’s normal behavior have an influence on the accuracy. Once the model is trained, it can be used for predictions based on real-time data collected through the mobile phone. ML techniques are well documented in literature [6,9,12] and are not further explored here.
2.4 Data Post-processing
Characteristic behavioral changes can be associated with symptoms based on the classification scheme from behavioral patterns. In the Susceptible, Infectious, Recovered or SIR model especially in the S(usceptible) to I(nfectious) transition phase user behavior changes [3] and can thus be used to improve prediction accuracy.
To analyze the temporal relationship of the behavior, Granger causality analysis has been used. The traditional linear Granger test has been widely used to examine the linear causality among several time series in bivariate settings as well as multivariate settings [14]. It is used to determine if one time series has predictive information for another. For a behavioral pattern a time series can be for instance the coordinates of places a patient frequently visited during the training phase, for instance the coordinates of his work place or favorite café. The second time series are the coordinates of places he visits over time during the testing phase.
The original Granger tests examined the linear causality among several time series in a bivariate and multivariate setting. However many real world applications are nonlinear and extensions have been developed [14, 16] to overcome this constraint. Recently the Phase Slope Index (PSI) has been preferred over Granger causality in some studies [3], [17]. PSI is a recently proposed spectral estimation method designed to measure temporal information flux between time series signals [3]. It is based on the assumption that the information flux between two signals can be estimated using the phase slope of the cross-spectrum of the signals. Independent noise mixing does not affect the complex part of the coherency between multivariate spectra, and hence PSI is considered more noise immune than Granger analysis [3]. The Phase Slope Index is defined as:
Where Cij is the complex coherency, δf =1/T is the frequency resolution, and ℑ (·) denotes taking the imaginary part. PSI has been used to validate causal links between time series of symptom days where participants showed stress and depression symptoms [2].
3 Challenges and Ethical Issues
Behavioral patterns are highly personal and vary from individual to individual. Behavioral patterns of introverts, persons lacking social skills, lethargic or isolated persons show smaller variations when sick than active, sociable persons. Training and predictive models have to be enough granular to capture and detect deviant behavior of patients with a big variety of different behavioral patterns. There are many reasons why behavioral patterns change. Students before examinations spend more time studying and are less engaged in physical activities. Someone in a new relationship might change his behavioral patterns. The challenge of correctly classifying behavior and avoiding false positives based on misinterpretation like “work at home” interpreted as deviant behavior has to be addressed by any real-world application. Recording and analyzing the behavioral patterns of patients in real-time raises serious privacy issues. It represents a high level of surveillance where every movement and conversation is logged for analysis. There are also security issues. Announcing a person’s location to the world can tip off burglars or stalkers.
4 Conclusions
While reality mining on mobile phones in the health care sector is still in its infancy, there are already promising applications. Modern societies face the challenge of caring for their aging population. Applications of reality mining using mobile phones might help elderly people, people with disabilities or diseases like Alzheimer’s living safer and more independently and reduce health care costs. But there seems to be no boundary for further applications on the individual level as well as on the public health level. Reality mining has already been used to measure social interactions or movement patterns of populations to determine the spread of infectious diseases and studies have buttressed the effectiveness of cell phones for early detection of outbreaks of epidemics [1,2]. There are already projects studying the spread of diseases in Africa [7]. Our findings suggest that it might be possible to answer such questions in the near future and to begin planning how to influence the development of even greater health-sensing capabilities in smartphones [2].
Lastly reality mining has shown that humans are more predictable than believed and that it is thus possible to reveal the identity of a person even if the mobile phone data is anonymized. More research in anonymizing behavioral patterns in reality mining would be highly desirable especially when used in the eHealth area.
References
1. Chronis, I., Madan, A., Pentland, A.: SocialCircuits: the art of using mobile phones for modeling personal interactions. In: Proceedings of the ICMI-MLMI 2009 Workshop on Multimodal Sensor-Based Systems and Mobile Phones for Social Computing, Cambridge, Massachusetts, pp. 1–4 (2009)
2. Madan, A., Cebrian, M., Moturu, S., Farrahi, K., Pentland, A.: Sensing the “Health State” of a Community. IEEE Pervasive Computing 11(4), 36–45 (2012)
3. Madan, A., Cebrian, M., Lazer, D., Pentland, A.: Social sensing for epidemiological behavior change. In: Proceedings of the 12th ACM International Conference on Ubiquitous Computing, Copenhagen, Denmark, pp. 291–300 (2010)
4. Gundecha, P., Liu, H.: Mining Social Media: A Brief Introduction. Informs 9, 1–17 (2012)
5. Pentland, A.: Automatic mapping and modeling of human networks. Physica A: Statistical Mechanics and its Applications 378(1), 59–67 (2007)
6. Wlodarczak, P., Soar, J., Ally, M.: What the future holds for Social Media data analysis. World Academy of Science, Engineering and Technology 9(1), 545 (2015)
7. Big Data Gets Personal. MIT Technology Review 16(4) (2013)
8. Tretyakov, K.: Machine learning techniques in spam filtering. In: U. o. T. Institute of Computer Science (ed.) Data Mining Problem-Oriented Seminar, p. 19 (2004)
9. Witten, I.H., Frank, E., Hall, M.A.: Data Mining, 3rd edn. Elsevier, Burlington (2011)
10. Sookhanaphibarn, K., Thawonmas, R., Rinaldo, F., Chen, K.-T.: Spatiotemporal analysis in virtual environments using eigenbehaviors. In: Proceedings of the 7th International Conference on Advances in Computer Entertainment Technology, Taipei, Taiwan, pp. 62–65 (2010)
11. Eagle, N., Pentland, A.: Eigenbehaviors: identifying structure in routine. Behavioral Ecology and Sociobiology 63(7), 1057–1066 (2009)
12. Liu, B.: Sentiment Analysis and Opinion Mining. Morgan & Claypool (2012)
13. Liu, B.: Web Data Mining: Exploring Hyperlinks, Contents, and Usage Data, 2nd edn. Springer, Heidelberg (2011)
14. Bai, Z., Wong, W.-K., Zhang, B.: Multivariate linear and nonlinear causality tests. Mathematics and Computers in Simulation 81(1), 5–17 (2010)
15. Diks, C., Panchenko, V.: A new statistic and practical guidelines for nonparametric Granger causality testing. Journal of Economic Dynamics and Control 30(9–10), 1647–1669 (2006)
16. Hiemstra, C., Jones, J.D.: Testing for Linear and Nonlinear Granger Causality in the Stock Price- Volume Relation. The Journal of Finance 49(5), 1639–1664 (1994)
17. Nolte, G., Ziehe, A., Krämer, N., Poupescu, F., Müller, K.-R.: Comparison of granger causality and phase slope index. In: NIPS 2008 Workshop on Causality, Canada (2008)
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
zd.nie@siat.ac.cn
Abstract. IEEE 802.15.4 and IEEE 802.15.6 are two kinds of wireless area network standards for short range communication applications. IEEE 802.15.4 is proposed for Wireless Person Area Network (WPAN) that provides low data rate, low power, and low cost applications in a short range. Meanwhile, IEEE 802.15.6 is the first international Wireless Body Area Network (WBAN) standard, which distributes nodes on or inside a human body, also operates in low power and short range, mainly provides real-time monitoring and human physiological data to judge the human physiological condition. In view of many similarities in both two standards, we analyzed the two standards mainly from the MAC frame format, MAC access mechanisms in this paper. In addition, some discussions of the differences of applications in the two standards were illustrated.
Keywords: WBAN · MAC · IEEE 802.15.4 · IEEE 802.15.6
1 Introduction
Nowadays, with the advancement of microelectronics technology, it becomes one of the most leading force to improve human existence and lifestyle through combing computer technology and communication technology. In this instance, the Wireless Sensor Networks (WSNs) with short distance, strong mobility, and high transmission rate is becoming more and more necessary and popular.
A WSN is composed of a large number of sensor nodes, and these nodes communicate with each other by self-organization and multi-hop [1]. A WSN is mainly used for monitoring physical or environmental conditions, such as temperature, sound, pressure, and to cooperatively pass their data through the network to a main location [2]. Earlier, several wireless communication standards have been formulated [3], such as the IEEE 802.11 [4], IEEE 802.15.1 [5], IEEE 802.15.4 [6] standards. However, these standards are not suitable for WBAN applications. The power consumption of 802.11 Wireless Local Area Network (WLAN) is too high to satisfy the wear WBAN requirements with a low power. In addition, the number of auxiliary nodes in IEEE 802.15.1 are limited. IEEE 802.15.4 is widely used in industrial sensors, smart grids
X. Yin et al. (Eds.): HIS 2015, LNCS 9085, pp. 7–16, 2015.
DOI: 10.1007/978-3-319-19156-0_2
and other areas of IOT (Internet of Things) [7], but it is not enough to support high data rate applications (data rate > 250 Kbps). In order to develop a low power consumption communication standard which is suitable for WBAN application [8], IEEE 802 established a task group for the standardization of WBAN called IEEE 802.15.6 in November 2007 [9]. WBAN is centered on the human body, which is composed of network elements (including personal terminal, independent nodes that are situated in the clothes, on the body or under the skin of a person, and communication equipment near human body within 3~5m) and so on [10]. WBANs provide unconstrained freedom of movement for patients suffering from chronic diseases, such as diabetes, heart disease [11]. The advantage is that a patient doesn’t have to stay in bed, but can move everywhere freely, which improves the quality of life for patients and reduces hospital costs. In February 2012, the first version of IEEE 802.15.6-2012 was published. IEEE 802.15.6 is a standard for short-range, wireless communications in the vicinity of, or inside, a human body (but not limited to humans) [9]. It defines a Medium Access Control (MAC) layer that works at lower sublayer of the data link layer of the OSI (Open System Interconnection) model [12], and offers unicast, multicast, or broadcast communication service. Unfortunately, more protocol details are hidden in current version of IEEE 802.15.6 standard, it is a better way to design a new WBAN system based on IEEE 802.15.4 standard, which is a mature protocol and has been applied in many fields. So far many research groups have studied the key issues of IEEE 802.15.4 and IEEE 802.15.6, the IEEE 802.15.6 MAC, PHY (Physical Layer), and security specifications were reviewed in [3]; the IEEE 802.15.4 security framework for WBAN was analyzed in [13]; the 802.15.4 MAC protocols for WBANs was introduced in [14]. However, few study was conducted to compare the similarities and differences between IEEE 802.15.4 and IEEE 802.15.6.
Analyzing and comparing 802.15.6 with 802.15.4 would help developers choose the better communication protocol to design new application systems and propose some approaches to optimize the 802.15.6 standard. In order to introduce the differences between the two standards, we discussed the MAC sublayer starting from the MAC format and access mechanisms, because the MAC sublayer plays an important role in providing guarantee for the reliable communication between SSCS (ServiceSpecific Convergence Sublayer) and PHY. The MAC sublayer concludes the MAC frame format, access mechanism and security services. The MAC frame format is used to indicate the frame types with different functions. The MAC access mechanism provides guarantee for the reliable communication. The security services make sure information safety. The MAC frame format and access mechanisms occupy an important position in the MAC sublayer, so we analyzed the two standards mainly from the MAC frame format, MAC access mechanisms in this paper.
The rest of the paper was organized into five sections. Section 2 presented the differences of the MAC frame format in the two standards. Section 3 introduced the different MAC access mechanisms between the two standards. A discussion of the differences of applications in the two standards was illustrated in section 4. The final section concluded our work.
2 Mac Frame Format of the Two Standards
A MAC frame is a sequence of fields in a specific order. The MAC frame format is composed of a MAC Header, a MAC Payload, and FCS both in IEEE 802.15.4 and IEEE 802.15.6. As depicted in Table 1.
Table 1. General MAC frame format
Octets: variablevariable 2
MAC Header MAC PayloadFCS
If a device wants to transmit data to other devices in IEEE 802.15.4, it should contains a MAC Header with at least 9 octets length which is longer than those with 7 octets length in IEEE 802.15.6. These may result in many difficulties during the frames transmission and reception, such as increasing the burden of transceivers. In addition, the data rate will decrease and the transmission power will increase.
2.1 Frame Control field
Besides the intuitionistic difference of the length of MAC Header, there are many similarities and differences between the two frame control fields, as described in Table 2 and Table 3.
2. Format of the Frame Control field in IEEE 802.15.4
Bits:0-23 4 5 6 7-9 10-11 12-1314-15
The Frame Type subfield of IEEE 802.15.4 is 3 bits in length and shall be set to one of the non-reserved values. 0b000, 0b001, 0b010, 0b011 respectively denote the beacon frame, data frame, response frame, MAC command frame, other values were reserved. The IEEE 802.15.6 describe the Frame Type by using not only Frame Type subfield but also Frame Subtype subfield. 0b00,0b01,0b10 of the Frame Type subfield represent management frame, control frame and data frame respectively, 0b11 is reserved. On the other hand, the Frame Subtype subfield refines the frame type with 4 bits data, which is helpful to classify different frames. For example, if a frame field carries
Table
Table 3. Frame Control format of IEEE 802.15.6
security association information, it must be a management frame not a MAC command frame. In other words, a combination of Frame Type subfield and Frame Subtype subfield is more efficient than an independent use of Frame Type subfield. In addition, the IEEE 802.15.6 defines an UP (User Priority) to decrease collision possibility.
2.2
MAC Frame Body
In IEEE 802.15.4, the Frame Payload field has a variable length and contains information specific to individual frame types. The PHR (PHY header) frame length field identifies the length of the MAC frame, it is a byte long and the MSB of the PHR frame length field is not valid, so the length of the MAC frame can’t exceed 127 bytes, which is not suitable to use the RTS/CTS mechanism. Nor is IEEE 802.15.6 with a variable length from 0 to 255 bytes of MAC Frame body. Because a RTS package with 20 bytes in length could account for about 20% of the MAC Frame body in IEEE 802.15.4 and 10% of those in IEEE 802.15.6 respectively, which lead to extra energy consumption, meanwhile, the RTS/CTS mechanism couldn’t effectively restrain hidden conflicts [15]. In IEEE 802.15.6, when MAC Frame body has a nonzero length, it contains 1-bit Low-Order Security Sequence Number, a variable length of Frame Payload (Do not exceed pMaxFrameBodyLength) and 4-bit MIC (Message Integrity Code). The Low-Order Security Sequence Number field carries message freshness information required for nonce construction and relay detection. In addition, the last 32-bit MIC (Message Integrity Code) carries information about the authenticity and integrity of the current frame.
3 MAC Access Mechanism
As shown in in Figure 1, the IEEE 802.15.4 communication mode is represented. In IEEE 802.15.4, the beacon-enabled with superframe uses a slotted ALOHA or slotted CSMA/CA in CAP and GTSs in CFP to exchange information between the coordinator and devices. In addition, unslotted CSMA/CA mechanism is used by nonbeacon without superframe.
Fig. 1. IEEE 802.15.4 communication mode
Analysis and Comparison of the IEEE 802.15.4 and 802.15.6 Wireless Standards 11
Similarly, the IEEE 802.15.6 also employ the slotted ALOHA and CSMA/CA mechanism, what’s more, there are two other protocols named improvised and unscheduled access mechanism and scheduled and scheduled-polling access mechanisms.
3.1
Random Access Mechanism
In EAP, RAP, and CAP periods of beacon mode with superframe boundaries of IEEE 802.15.6, as shown in Figure 2, the hub may employ either a slotted ALOHA or CSMA/CA protocol, depending on the PHY. To send data type frames of the highest UPs (User Priorities) based on CSMA/CA, a hub or a node may combine EAP1 and RAP1 as a single EAP1 and EAP2 and RAP2 as a single EAP2, so as to allow continual invocation of CSMA/CA and improve channel utilization. When using slotted ALOHA for high-priority traffic, RAP1 and RAP2 are replaced by another EAP1 and EAP2 respectively but not a continuation EAP1 and EAP2, due to the time slotted attribute of slotted ALOHA access. B EAP1RAP1MAPEAP2RAP2MAP
CAP Superframe (beacon period)
Fig. 2. Beacon mode with superframe boundaries in IEEE 802.15.6
In a slotted ALOHA protocol, the nodes access the channel using predefined UPs, as given in Table 4.
Table 4. Bounds for slotted-ALOHA and CSMA/CA protocols
User Priorities Slotted –ALOHACSMA/CA
Initially, the CP (Collision Probability) is selected according to the UPs. If Z≤CP, where Z is equal to a random number in the interval [0-1], the node obtains a contended allocation in the current ALOHA slot, during which data frames transmission occur. When the transmission is fail, the CP remains the same if the number of failures are odd or be cut in half if the number of failures are even.
The IEEE 802.15.4 protocol defines two versions of the CSMA/CA mechanism: slotted CSMA/CA mechanism for beacon mode with superframe and unslotted CSMA/CA mechanism for non-beacon network. In both cases, the algorithm is implemented using units of time called backoff periods, which is equal to aUnitBackoffPeriod symbols. The CSMA/CA algorithm is controlled by three variables: NB (Number of Backoffs), CW (Content Window) and BE (Back off Exponent). Where NB is initialized to zero and the maximum value is 4. CW is decreased using units of backoff, the default value is 2 and the maximum is 31. BE is related to how many backoff periods a device shall wait before attempting to assess the channel and the scope of BE is 0~5, the default value is 3. The whole CSMA-CA algorithm is illustrated in Figure 3.
CSMA/CA Slotted?
NB=0,CW=2
BE=lesser of (2,macMinBE) Y Y
Battery life extension? BE=macMinBE
NB> macMaxCSMA Backoff? Failure N Y (4) (3) (2) N CW=CW-1 CW=0?
Located backoff Peirod boundary
Delay for Random (2^BE-1)unit Backoff period
Perform CCA on Backoff period boundary
Channel idle?
CW=2,NB=NB+1, BE=min(BE+1,macMaxBE)
Y N (5) N
NB=0,CW=2
Delay for Random (2^BE-1) unit Backoff period
Perform CCA
Channel idle?
NB=NB+1, BE=min(BE+1,macMaxBE)
NB> macMaxCSMAB ackoff?
Failure Y N
Fig. 3. The CSMA/CA algorithm in IEEE 802.15.4
In IEEE 802.15.6, the node initially sets BC (Backoff Counter) to a random inter that is uniformly distributed over the interval [1, CW], where CW Є ( , ).
As shown in Table 4, the values of and are selected according to the
UPs. Before implementing the CSMA/CA algorithm, the CSMA slot boundary and pSIFS should be located, which each idle CSMA slot is equal to pCSMASlotLength and the default value of pSIFS is 75 μs. The m is the times of the node had failed consecutively. It is important to note that if double of the CW exceeds the , then the CW is . Figure 4 shows an example of the CSMA/CA algorithm.
Comparing the two CSMA/CA protocols in IEEE 802.15.4 and IEEE 802.15.6, it is found that there are some similarities and differences between them. First of all, since they are both CSMA/CA, so the node needs to detect the channel by using CCA (Clear Channel Access) before transmitting frames. CW is both used to implement backoff algorithm which is not the same in using. In IEEE 802.15.4, BE is related to how many backoff periods a device shall wait before attempting to assess the channel and the BEth backoff is randomly chosen from {0,1,…, 2 -1}, this is done to reduce the probability of the same backoff period for different nodes. However, owing to the less nodes, the shorter distance, the faster rate, the IEEE 802.15.6 defines UPs to decrease collision possibility. Different UPs mean different CW and BC, additionally, smaller CW and BC lead to low latency and higher channel utilization. All these designs are adopt to IEEE 802.15.6 network topology.
CSMA/CA
Located CSMA slot boundary
Delay for pSIFS BC unlocked
CCA on CSMA slot boundary
Fig. 4. The CSMA/CA algorithm in IEEE
3.2 Improvised and Unscheduled Access Mechanism
Besides the slotted ALOHA protocol and the CSMA/CA mechanism both in IEEE 802.15.4 and IEEE 802.15.6, there are another two access mechanisms in IEEE 802.15.6. The hub may use improvised access to send poll or post commands without advance reservation in beacon or non-beacon modes with superframe boundaries. Figure 5 illustrates an example of immediate polled allocations.
5. Immediate polled allocations
The hub may also use an unscheduled access mechanism to obtain an unscheduled bilink allocation to apply for some specific condition, such as emergency communication service. In beacon or non-beacon modes with superframe, unscheduled bilink allocations may be 1-periodic, where frames transmission every superframe, or m-periodic, where frames transmission every m superframes. An m-periodic bilink allocation is helpful to reducing power consumption because nodes could sleep in m-periodic allocation.
3.3 Scheduled and Scheduled-Polling Access Mechanism
Unlike unscheduled allocation, a node and a hub may employ scheduled access to obtain scheduled uplink, downlink, and bilink allocations. In addition, the scheduled polling is used for polled and posted allocations. These allocations may be 1-periodic or m-periodic, but not the both in the same BAN. Figure 6 illustrates an example of scheduled 1-periodic allocations.
(superframe) n+1
8 9
8 9
(superframe)
period (superframe) ...
8 9
period (superframe)
Fig. 6. Scheduled one-periodic allocation
Fig.
4 A Discussion of Differences of Applications
The standards are described differ by which frequencies they used and the data rate and range they covered. According to Table 5, 802.15.6 (BAN) has a much shorter range than 802.15.4, which proves to be an advantage. Shorter range communication means lower power requirements. In addition, due to the short distance in the vicinity of, or inside, a human body only, it’s more secure than 802.15.4. The lower consumption of the IEEE 802.15.6 not only comes from the sleep mode as same as IEEE 802.15.4, but also the shorter distance and le ss interference. In addition, it enables equipment to be smaller and frequency reuse to be better. The data rate of 802.15.6 is up to 15.6 Mbps, which is much faster than 802.15.4 with a max data rate 250 kbps [16]. All of these are good for BAN since the design of original intention is to make it unobtrusive, you can put it in the clothes, attach or implant into the human body, such as wearable devices.
5. Comparison of the two wireless standards
MAC frame type
MAC access mechanism
beacon frame, data frame response frame and command frame
Slotted CSMA/CA
Unslotted CSMA/CA
management frame, control frame and data frame
CSMA/CA mechanism
Improvised and unscheduled access mechanism
Scheduled and scheduled-polling access mechanism
Data Rate (Max) 20kbps,40kbps/250kbps 15.6Mbps
Transmission Range 75m 3~5m
Applications
Low data rate, industrial sensors, smart grid Wearable devices
The major reason in increasing data rate is that the transmission medium is the human body, which with little interference. Coupling with new modulation techniques, making it possible that the data transfer rate of IEEE 802.15.6 is much higher than IEEE 802.15.4. In a word, IEEE 802.15.6 defines a new wireless communication technology for low power, high data rate, short range, high safety which is especially suitable for wearable device applications.
5 Conclusions
This paper presented the most significant features of comparison of MAC between IEEE 802.15.6 and IEEE 802.15.4 standard. An analysis of differences of MAC format and access mechanisms of the two standards were presented. At last, starting from the aspects of frequency, data rate and range, the superiority of 802.15.6 in BAN communication was discussed. We believed that this paper could be used to quickly
Table
understand the key feature of MAC sublayer of IEEE 802.15.4 and IEEE 802.15.6. Besides, it also helped you to develop the potential application of IEEE 802.15.6 on the basis of IEEE 802.15.4.
Acknowledgment. This study was financed partially by National Natural Science Foundation of China (Grant No.61403366), Shenzhen Basic Research Project Fund (JCYJ2014041711 3430695), the National 863 Program of China (Grant No. 2012AA02A604), the Next generation communication technology Major project of National S&T (Grant No. 2013ZX03005013).
References
1. Heragu, A., Ruffieux, D., Enz, C.: The design of ultralow-power MEMS-based radio for WSN and WBAN. In: Frequency References, Power Management for SoC, and Smart Wireless Interfaces, ed , pp. 265-280. Springer (2014)
2. Rawat, P., Singh, K.D., Chaouchi, H., Bonnin, J.M.: Wireless sensor networks: a survey on recent developments and potential synergies. The Journal of Supercomputing 68, 1–48 (2014)
3. Ullah, S., Mohaisen, M., Alnuem, M.A.: A review of ieee 802.15. 6 mac, phy, and security specifications. International Journal of Distributed Sensor Networks (2013)
4. T. W. G. f. W.: Standards. IEEE WLAN (2012). http://www.ieee802.org/11/
5. IEEE WPAN Task Group 1 (2012). http://www.ieee802.org/15/pub/TG1.html
6. IEEEStd.802.15.4:WirelessMedium Access Control(MAC) and Physical Layer (PHY) Specifications for Low Data Rate Wireless. IEEE Std 802.15.4™-2006 (2006)
7. Gubbi, J., Buyya, R., Marusic, S., Palaniswami, M.: Internet of Things (IoT): A vision, architectural elements, and future directions. Future Generation Computer Systems 29, 1645–1660 (2013)
8. Nie, Z.D., Ma, J.J., Ivanov, K., Wang, L.: An investigation on dynamic human body communication channel characteristics at 45MHz in different surrounding environments. In: Antennas and Wireless Propagation Letters, p. 1. IEEE (2014)
9. IEEE Standard for Local and metropolitan area networks - Part 15.6: Wireless Body Area Networks, IEEE Std 802.15.6-2012, pp. 1-271 (2012)
10. Chávez-Santiago, R., Khaleghi, A., Balasingham, I., Ramstad, T. A.: Architecture of an ultra wideband wireless body area network for medical applications. In: 2nd International Symposium on Applied Sciences in Biomedical and Communication Technologies, ISABEL 2009, pp. 1-6 (2009)
11. Kim, T.-H., Kim, Y.-H.: Human effect exposed to UWB signal for WBAN application. Journal of Electromagnetic Waves and Applications, 1-15 (2014)
12. Kumar, M. G., Roy, K.S.: Zigbee Based Indoor Campus Inventory Tracking Using Rfid Module
13. Saleem, S., Ullah, S., Kwak, K.S.: A study of IEEE 802.15. 4 security framework for wireless body area networks. Sensors 11, 1383–1395 (2011)
14. Ullah, S., Shen, B., Islam, S.M., Khan, P., Saleem, S., Kwak, K.S.: A study of MAC protocols for WBANs. Sensors (Basel) 10, 128–145 (2010)
15. Barroca, N., Borges, L. M., Velez, F.J., Chatzimisios, P.: IEEE 802.15. 4 MAC layer performance enhancement by employing RTS/CTS combined with packet concatenation. In: 2014 IEEE International Conference on Communications (ICC), pp. 466-471 (2014)
16. Nie, Z., Ma, J., Li, Z., Chen, H., Wang, L.: Dynamic propagation channel characterization and modeling for human body communication. Sensors 12, 17569–17587 (2012)
Another random document with no related content on Scribd:
producing a lovely colored light. These glow-worms are somewhat rare in this country, but are common in England, where our fire-fly is unknown. The glow-worms of tropical countries are as large as good-sized caterpillars, and give out a light of corresponding size.
Some of the beetles of tropical countries are much more radiant than the glow-worms. They have a bright ring around their bodies, which sheds such a light that it is said that the negroes use them for lamps. These negroes we are told by travelers make small round cages of thin wooden slats placed near each other, and closed at the top and bottom. In these they put several beetles, and thus light up their rooms free of cost.
THE GREAT LANTERN FLY
But the most splendidly illuminated insect in the world is the great lantern fly. Its monstrous head is a blaze of light. When it flies through the air it is like a streamer of fire. When it alights upon a leaf, with its beautiful wings outstretched, and its head gleaming with starlike rays, that light up everything for some little distance, it is a gorgeous sight indeed.
But, although this insect is so bright-headed it is not so wise as some others whose heads are not brilliant at all. Perhaps this head is
like some shops we have seen, where pretty much all the furniture is on the outside.
OWLS ON A FROLIC.
THE OWLS UPSETTING THE LAMP.
The owls are abroad on a mad carouse, Waking the echoes far and wide; They whirl in a crowd through the ruined church,
Or up to the belfry glide.
The little screech-owl makes a horrid din; While the great white owl looks wise; And the horned owl nods his head, and blinks; As around the lamp he flies.
The lamp is a cup, half filled with oil, That swings from a broken beam; And, over the traveler sleeping below, It throws but a dusky gleam.
The owls have no fear of the burning wick— ’Tis only a cotton loop— They’re after the oil in the swinging cup, And down on its brim they swoop.
The weary traveler, sound asleep, Hears naught of the noise o’erhead, A rickety chair as a bedstead serves, His overcoat is his bed.
With the sweep of the wings the lamp upsets, While the gurgling oil o’erflows With a drip, and a rush, on the great owl’s tail, A splash on the traveler’s nose.
He’s up in a trice, and, seizing a broom, He arms himself for a fight. But all is still in the ruined church; For the owls are out—and his light.
COMMON AND UNCOMMON SPONGES.
A FINE SPONGE.
They are all wonderful enough, no matter how common they may be. It takes thousands and thousands of minute creatures, to make a sponge, and these creatures are so little understood that about all we know of them is that they must belong to the very lowest order of
animal life, and that they do build sponges. That is not much to know, but it is not long since the sponge was first known to be an animal production at all, and our scientific men may yet find out something more definite about these curious little architects. Perhaps they may have lately found out something, and I have not heard of it. This would be the least wonderful thing about sponges.
The ordinary form of the sponge is familiar to nearly everybody who has ever been washed, and this picture gives a good idea of a fine large one, as it is found growing at the bottom of the sea. I say growing, because it seems to be growing there, like a vegetable. But it does not grow, in the ordinary sense of the word, any more than a wasp’s nest grows.
But there are sponges with which we are not at all familiar, and which are curious, apart from the manner of their construction. Such a one is the sponge called the “Cup of Neptune.”
This is several feet high, and is formed like a great goblet. It would make a very good cup for Neptune, if he drank brandy or rum, for it would soak up all that he poured into it, and he could not get a drop, unless he squeezed his cup pretty hard—and even then the liquor might all run out of the bottom.
As a rule, civilized and well educated people are more easily surprised and astonished at uncommon and wonderful works of nature than uncivilized or ignorant people, for the latter do not know enough to be astonished. They see nothing strange in the development of a plant from its seed—nothing grand in a high mountain, nor anything very beautiful in a flower. They look at these things as a child looks at his hand. The hand is a very curiously constructed instrument, full of intricate mechanism, but the child does not know or think of that. It is not until he grows older and his mind is cultured that he appreciates the wonderful construction and the varied action of his hand.
So it is with savages. They do not comprehend that many strange works of nature are worthy of admiration, and they take it for granted that things are as they are because they ought to be, just as they think of their own bodies, if they think of them at all.
THE CUP OF
But this great goblet-like sponge is strange enough to astonish even a savage.
NEPTUNE.
MAGHAR’S LEAP.
It chanced upon a time, a very great many years ago, while fairies and magicians still dwelt upon earth, that a youth and maiden— brother and sister—were walking in a forest, talking about their recent misfortunes, and laying plans for their future. The youth was clad in armor, according to the warlike fashion of those times. But he had under his arm a book, which was not in accordance with the fashion of those times. The maiden wore a dress of some coarse woolen stuff; and, in her hands she held a sheet of parchment, and a pen.
Suddenly there broke into their quiet talk the sound of clashing arms, and the mad plunging of horses. Sybil, the maiden, stopped terrified.
“Oh!” she cried, “it is the noise of battle! Too well I know those sounds. Let us go quickly back!”
“Let us go forward a little way,” said Maghar, the youth, “to yon opening in the woods. Or, stop here, if you fear, and I will go alone and look out.”
“No,” said Sybil, “if you go I will follow.”
Together they looked out upon the open plain. Two hostile armies had met unexpectedly, and a fierce conflict had commenced.
“Alas!” said Sybil, shuddering. “There are the savage infidels that laid waste our home!”
“Yes,” said Maghar, excitedly, “and here, on this side, are our countrymen, and neighbors! I must bear a hand in this fight!”
“And leave me alone!” cried Sybil. “I have only you left! Your single arm will not count for much in a battle!”
“It would be a shame to me,” said Maghar, “to sneak off, like a coward, and leave our friends and Christian soldiers, when their
forces are few, and every warrior counts. Have I not my armor? I shall find shield and spear on the battle-field on some poor fellow who has already fallen in the fray. Do not fear, sister! Go back to Christern’s cottage. There you will be safe; and I will return in a few hours.”
So saying, he led Sybil back into the forest to the path leading to Christern’s cottage; gave the book into her hands; and, kissing her good-bye, he ran out of the woods as fast as the weight of his armor would allow.
But Sybil did not return to the cottage. She was too anxious about her brother; and, going to the entrance of the wood, she crouched among the trees, where she was hidden from view, and watched the progress of the fight. She was ready to fly if the tide of battle brought the armies too near. But they seemed to be gradually moving away from her. She soon singled out her brother. He had secured a spear
SYBIL’S WATCH.
and shield, and mounted a riderless horse. In a few minutes he was lost in the throng, and she saw him no more.
Her mind was filled with sad forebodings. This Infidel army had invaded the country, and laid it waste; had killed her parents, and overthrown, and utterly ruined the beautiful castle that had been her home. A few things had been saved by old Christern, a much loved servant of the family, and these constituted the property of Maghar and Sybil. Old Christern’s cottage, in the depths of the forest, was the refuge of the orphans. There they had lived for several weeks, and no way of retrieving their fortunes seemed open to them. Maghar was a fine scholar. His father had had him taught to read his own language and Latin, and to write a very beautiful hand. That was the extent of his knowledge; and it was a great deal at a time when very few of the richest people knew their letters.
And now, in their poverty, there seemed to be very little use for his learning. Nobody cared anything about it. He might copy manuscript for some learned man, and get a living this way, for printing and paper had not then been invented; and all books were written on parchment. But Maghar had a contempt for a clerk, as he called a copyist, and did not fancy this method of supporting his sister and himself. Nevertheless, the two were that day on their way to the abode of a great and learned man to see if he wished anything of this kind done; and if he would buy their only book—a Latin volume, written on parchment, and beautifully illuminated and bound in wooden covers.
Sybil went over these things in her mind as she watched the battle, trying, in vain, to distinguish the form of her brother. She soon saw to her dismay, that the Infidel forces had turned the flank of the Christian army, and that the ranks of the latter were broken, and they were retreating, closely followed by their enemies. She stood up now, and strained her eyes to watch them until they had all disappeared over the crest of a hill. Then she sadly returned to Christern’s cottage to tell the old man of this new and terrible misfortune.
Days passed away, and Maghar did not return. Christern learned that the Christian army was broken, and the soldiers scattered. Some had returned to their homes. The wounded were cared for among their friends. The dead were buried. But Maghar was with none of these. No one could tell anything about him, except that he had fought bravely.
Then Sybil determined to seek out the great and learned man to whom Maghar had intended to offer his services as copyist. She was somewhat afraid of him, for he was known to be a powerful magician. But he could, no doubt, tell her the fate of Maghar, and she would try to overcome her fears.
She took off the coarse peasant’s dress she had been wearing, and arrayed herself in her best robe of fine white cashmere, which was one of the things that Christern had managed to save. She loosened her beautiful hair, which fell nearly to her feet. This last she did to show the deep sorrow she was in. She also took with her the Latin volume, as a present, to propitiate the powerful magician.
The great man lived in the simplest manner in a rocky cavern. Sybil found him outside his dwelling, seated on a mossy stone, sorting some plants that lay in his lap. He did not look up as she approached, and she had a good opportunity to study his countenance, which was so sweet and gentle that her fear of him vanished; and she came forward quite boldly, greeted him, and presented her book.
But the magician waved the volume aside. “I know why you seek me, sister of Maghar,” he said, kindly.
“Oh, can you tell me aught of my brother?” cried Sybil.
“I know not where he is. The oracles would not enlighten me without your presence. Come into my dwelling, and we will consult them.”
So saying he conducted her into his cave through a low, dark passage way. Great was Sybil’s astonishment when she found herself in a vast room, with a lofty ceiling. Around the circular walls was a continuous row of lamps, kept constantly burning. Their light
was reflected from myriads of stalactites that hung from the roof, glowing with all the colors of the rainbow, making the rough, rocky chamber as brilliant and gorgeous as a fairy palace. In the centre of the room stood a brazier, filled with burning coals, and near it, a large iron harp, with silver strings, and a sort of cupboard, made of iron. A few rough couches were scattered around. And this was all the furniture the room contained.
The magician invited Sybil to take a seat. He then proceeded to place on his head a crown, woven of vines of magical virtues. He took from the cupboard some singular-looking vessels, and mixed in them various powders and liquids. Then, pouring all their contents into a copper pot, he placed it on the coals, seated himself on a stone near it, drew his harp in front of him, and motioned to Sybil to stand before it. He looked so pleasantly upon her she did not feel afraid, but her heart beat fast, not knowing what fearful thing she might see.
She saw nothing whatever but the harp, and the old man; for, as soon as the clouds of fragrant white smoke that poured out from the brazier, had completely enveloped the two, the magician swept his fingers over his harp, and began to sing. Then Sybil forgot everything else, for his chant was of Maghar.
SYBIL AND THE MAGICIAN.
He sang of the great deeds Maghar had done in the battle, and how he had made himself famous. He was the last prisoner taken by the Infidels; and was now confined in a castle several leagues distant. The Infidel army was there encamped. They would like to slay Maghar outright, but were afraid of the vengeance of the Christian armies near them if they murdered a man held in such esteem. He was at present undisturbed, but the probability was that,
after a time, they would decide to starve him to death, and give out word that he had died from sickness. His sister had thus a little time in which to work to save him.
Here the song ended, and the weeping girl begged the great magician to save her brother. This he said was not in his power. She must find a good fairy, and make it her friend. The small creature could get into the castle, see her brother, and, together, they could devise a way of escape. He might, perhaps, be able to help them then. He told her what roads to follow to reach the castle; and, assuring her that such a good girl would surely find a good fairy to assist her in her trouble, he dismissed her with his blessing.
That very day Christern and Sybil set out for the castle. They reached the place after three days’ journey. They told no one what their errand was in that part of the country; and there were so many homeless people in the land that their appearance excited no surprise. Christern soon found employment among the wood-cutters, and fitted up a deserted hut as a temporary dwelling.
But though they could, every day, look upon the walls of the castle in which Maghar was confined, they seemed no nearer to him than before. He was in the hands of the cruel infidels, and where were there any fairies? There were plenty in that part of the country, the wood-cutters said, which, at first, was encouraging. But, on inquiry, it turned out that not one of them had ever seen a fairy, or knew anybody who ever had seen one. Sybil was in despair as the days went by, and she blamed her friend, the magician, that he had given her no help, after all.
She often walked through the woods, near nightfall, to meet Christern. One evening, as the two were returning together to their hut, they saw a large wild boar approaching, followed by several young ones. As this creature is very savage when it has its young to defend, Christern and Sybil thought it wise to step aside among the trees, and leave the path to the boar and its interesting family After these had passed they continued their way, but had not gone far when they saw a young boar lying in the path. Christern stooped over to examine it.
THE BOAR FAMILY.
“It got in with that litter,” said he, “and did not belong to it, so the old boar has gored it badly. But it is not dead. I’ll take it home, make a sty for it, and, if it lives, I’ll fatten it, and kill it when it is fit for eating.”
The wounded animal lifted an appealing glance to Sybil. Its eyes wore an almost human expression of suffering, and a most beseeching plea for help. The girl’s heart was touched.
“It is not badly hurt,” she said. “Its flesh is torn, but if I wash its wounds, and bind them up, and find a nice place in the woods, where I can make it comfortable, and feed it, it will get well. It is a free, wild creature, and must not be shut up in a close sty. Think of my dear brother shut up when he wants to be free!”
Christern thought Sybil’s plan a foolish one, but this last argument silenced him. He had not a word to say in reply. So the girl washed
off the blood from the boar’s wounds with her fine cambric handkerchief, which she then tore into strips to bind them up. She found, in a secluded place, a soft cushion of moss on which she laid him, and partly covered him with leaves to keep him warm. She then brought from the hut some of her own scanty supper, and gave it to the little boar.
After this she visited her patient two or three times a day, nursing and feeding him. But, on the afternoon of the fourth day, he had disappeared, and Sybil returned to the hut feeling quite lonely at the loss of the little creature that had been so glad to see her.
The next night, as Christern was returning late from his work, trudging slowly through the forest, with his lantern swinging in his hand, and his wallet slung over his back on the end of his walking stick, something brushed close by the old man’s ear with a buzzing of tiny wings.
“That dragon-fly is out late,” said the old man to himself.
Very soon the wings brushed by him again with a louder whizzing.
“It is a bat!” said the old man, shaking his head. “Shoo! shoo!” But the third time the whirring wings flew almost into his face.
“Good evening, old Christern!” said a tiny voice, such as might come from a humming-bird, if it could speak.
The startled old man stopped and flashed the light of his lantern around among the trees. And there, with wee wings outspread, was a fairy skimming through the air! Christern had never seen a fairy, but he knew this was one as soon as he saw him. And a jolly, rollicking fellow he was!
CHRISTERN AND THE FAIRY.
“You don’t know me, old fellow?” said the fairy.
Christern shook his head.
“Wanted to shut me up in a sty, and fatten me, eh? I wouldn’t be much of a mouthful now, would I? Don’t you wish you could get me?”
And the saucy fellow soared high up among the trees.
Christern nearly dropped his lantern in his astonishment. “You don’t ever mean to tell me that boars are fairies?” he said, at last.
“I mean to say nothing of the kind!” cried the fairy, indignantly. “Your horrid, beastly boars are no relations of ours, even! I’ll tell you how it was,” he said, coming nearer Christern, and speaking in a confidential tone. “Our fairies all have wings, and can fly, but there are other kinds without wings. Some of these are good, but some are bad, and they are full of spite against us because we are better off
than they I offended a tribe of these not long ago, and they had influence with a wicked old witch who changed me into a little boar. I was to remain in that shape for a week. She would have made the time longer, if she could. But they all thought I would be killed in that time. And so I should have been but for your Sybil. And there was another thing worse than death. If I was deprived of my liberty during that week, I could never again regain my natural shape. So, if you had put me in your sty, I would have been eaten up one of these days as a boar. From this awful fate your Sybil saved me. So I am doubly indebted to her, and I want to do something for her.”
“Oh, you are the good fairy, who is to save our Maghar!” cried the old man, joyfully.
Thereupon he related the whole sad story, and the fairy told him he would consult with his tribe that night; and, if he and Sybil would come to that spot on the following night he would let them know what could be done.
Sybil’s delight was unbounded. She now felt sure that her brother would be saved. But, nevertheless she accompanied Christern to the place of meeting, half fearing that the frisky fairy would play her some trick. But he was there, before them, and had dressed himself in his best suit of green in honor of the occasion.
As soon as they appeared he began chattering as fast as ever he could.
“We fairies have hit upon a splendid plan,” he said. “But there is no time to lose. Sybil, I have seen your brother, but he did not see me. I was at the castle this morning before cock-crow. I flew in through a loop-hole. Nobody saw me. It took me a long time to find out in what room your brother was kept, but, at last, I made it out. I intended to stay until I did. He is in a room, high up in the north tower. He has been pretty well, but now his jailers have begun the plan of starving him; and he will soon be too weak to save himself as we propose, which is the only way open to him. It requires steady nerves, and great courage. But do not weep, for we will save him, only it must be done speedily. Do you, Christern, be ready to go with me to the castle at break of day. Pretend you are a beggar. There are so many
of these you will pass unsuspected. I will point out to you a small postern door at the back of the castle, stay about that; and I will hide near it. I could slip inside easily enough, and tell Maghar what to do, but he does not know me, and would not trust me. So you must get inside the castle some way and see him. And, not only that, but you must get out again. And this is our plan for doing this. We fairies have three magical cocks. At a signal from me these cocks will appear on the crest of the hill at the back of the castle, and will sing a song. This will so astonish the sentinels that they will be thrown off their guard. I will then slip in through a loop-hole, unlock the postern door, and let you in. We will tell Maghar how he can escape. Then the cocks will appear again, and while the attention of the guards is distracted, we will get out of the castle. Remember now to be here at daybreak.”
And the fairy disappeared, much to Sybil’s regret, who had a hundred questions to ask him about her brother’s appearance, and treatment. He had not even told her what his plan was for her brother’s escape. But he did not come back, and she was obliged to be satisfied with the information she had.
The programme was carried out in every particular. Christern acted his part of beggar so well that he managed to get near the postern door, unsuspected, with the fairy snugly tucked into a fold of his ragged dress. On arriving at the place the fairy concealed himself in some vines. At the appointed signal three magnificent cocks appeared abreast on the top of the hill.
THREE MAGICAL COCKS.
The like of these cocks had never been seen in that country, and they immediately attracted the attention of everybody. But when they opened their mouths, and began to sing the words of a war song, the sentinels forgot everything, and deserted their posts to get as near the wonderful songsters as possible without alarming them.
