International Journal of Advances in Science and Technology, Vol. 3, No.

3, 2011

Perception of Beacon in localization of Wireless Sensor Networks

S.Arun , 2Krishnamoorthy and 3S.Ragavendran

Dept. of ECE,Veltech High Tech Engg. College ,Chennai [email protected]

Junior Research Fellow ,TCOE, Veltech Dr.Rr & SR Technical university,chennai

Asst.Professor, Veltech Dr.Rr & SR Technical university,chennai

Abstract
The Pervasive computing and sensor networks are emerging as key application drivers for wireless networks. Localization is of fundamental importance to pervasive computing and sensor networks because the former operate on devices based on their physical proximity and the latter names data and organize the network in terms of their physical location. Beacons (known-location nodes) are one key approach to achieving localization in wireless networks. In this paper we propose a Beacon Movement Detection Schemes such as Location based, Neighbor based, Signal strength binary and Signal strength real schemes for localization accuracy based on the. Automatic Monitoring, Identifying and Removal of such unreliable beacons from the localization Module with the following assumption that there are unnoticed changes of locations of some beacons in the system which affects localization accuracy. The effect of RSSI error in Signal Strength real scheme can be suppressed by individual diversity difference coefficient, distance difference coefficients and distance difference localization equation is defined based on analysis models of maximum likelihood estimation and RSSI, which improves localization

I.INTRODUCTION
Wireless sensor networks (WSNs) are essentially intended to observe spatial-temporal characteristics of the physical world. Locations of sensor nodes are fundamental to providing location stamps, locating and tracking objects, forming clusters, and facilitating routing, etc. However, a priori knowledge of locations is unavailable in large-scale and ad-hoc deployments, and a pureGPS (Global Positioning System) solution is viable only with costly GPS receivers and good satellite coverage. In a general scenario, only a few nodes (called anchors) are aware of their positions either through manual configuration or equipped with GPS receivers, and the others (called unknown nodes) have to estimate their positions by making use of the positions of anchors. Localization algorithms in WSNs are broadly divided into range-free approaches and range-based approaches. Range-free approaches normally rely on proximity, near far information or less accurate distance estimation to infer the locations of unknown nodes, and range based approaches require accurate distance or angle measurements to locate the unknown nodes .Both approaches must rely on the positions of anchor nodes and some measured/estimated parameters, and the localization accuracy depends on the accuracy of reference positions and relative parameters. In most localization systems, we assume that there are sets of beacon sensors which may or may not be aware of their locations and can periodically transmit/ receive short broadcast packets. By evaluating the distances, angles of arrival, or signal strengths of these broadcast packets. We can estimate the locations of objects by triangulation or pattern matching Under such an architecture, we observe that most existing works have an underlying assumption that beacons are always reliable. Based on this observation, this paper a new Beacon Movement Detection (BMD) . L ocalization system is unaware of this event. With unnoticed beacon movement events, the topology of the sensor network may be different from what it is supposed to be, and thus a localization algorithm may lose its accuracy In this paper, we assume that beacons are static under normal circumstances, but occasional beacon movement events are not unusual. This is true especially in a wireless sensor network. The BMD problem involves two issues. First, we need to monitor and identify the Beacons that change its location unexpectedly. Second, the result has to be given to the positioning Module to reduce the impact of Movement on localization accuracy. Based on this assumption, we propose four schemes. The first location-based (LB) scheme tries to calculate each beacons current location and compares the result with its predefined location to decide if it has been moved.

Special Issue I

Page 60 of 65

ISSN 2229 5216

International Journal of Advances in Science and Technology, Vol. 3, No.3, 2011 In the second neighbor-based (NB) scheme, beacons will keep track of their nearby beacons and report their observations to the BMD Module to determine if some beacons have been moved. In the third signal strength binary (SSB) scheme, the change of signal strengths of beacons will be exploited. In the last signal strength real (SSR) scheme,

Fig.1 System Model The BMD Module will collect the sum of reported signal strength changes of each beacon to make decisions. Note that only the first scheme assumes that the original locations of beacons are known in advance. The other three schemes do not assume any a priori knowledge on the original locations of beacons. In real time conditions, , signal strengths may be influenced by many factors, such as hardware difference, multipath propagation, and signal fading. To evaluate the proposed BMD schemes, we adopt a close-to-reality radio irregularity model to simulate the decay of signal strengths. This model has been shown to be able to reflect the propagation of radio signals, especially in indoor environments the proposed schemes shows their capability to improve the localization accuracy in events of beacon movement. Let us Assume a Sensing field with which a set of beacons H={H1,H2.Hn} is deployed for localization purposes. Depending on different schemes, the beacons locations may or may not be know in advance. Periodically, each beacon will broadcast a packet. To determine its own location, an object will collect packets from its neighboring beacons and send a signal strength vector S={s1,s2,..sn} to an external positioning Module, where si is the signal strength of the packet from Hi.If it cannot hear from Hi, we let si = smin, where smin denotes the minimum signal strength The positioning Module can then estimate the objects location based on S To solve the BMD problem, we will monitor each beacon from time to time. The content of an observation will depend on the corresponding BMD scheme. The BMD Module is capable of calculating a set BD. The result is then sent to the calibration algorithm in the positioning Module. The above Fig .1 illustrates our system model. Considering the following reasons, we define the tolerable region of each beacon Hi as the geographic area within which a slight movement of Hi is acceptable. First, radio signal tends to fluctuate from time to time. Second, slight movement of a Beacon should not change the signal strength much unless an obstacle is encountered. Third, ignoring the data of a slightly moved beacon in the location database will decrease the localization accuracy due to fewer beacons helping the localization procedure. So the slight movement of beacons is constrained by the tolerable regions. As a result, the unreliable set Dn only contains those beacons which are moved out of their tolerable regions. The sizes of tolerable regions are application dependent, which is beyond the scope of this work. For simplicity, tolerable regions are assumed to be circles centered at beacons of the same radius.

II.PREVIOUS WORK
There are two main approaches for localization: multi-lateration and pattern matching. Multilateration is a process of finding the location of an object based on measuring the distances or angles of three or more signal sources at known coordinates. A special case of multi-lateration is triangulation, ultrasonic sensors are used to estimate the location and orientation of a mobile device. In a distributed positioning system called Ad Hoc Localization System is proposed, where some beacons are aware of their own locations. The former are used to determine the positions of the latter. All the above systems require special hardware to support localization. Localization, using pattern-matching techniques is based on the localization task that can be achieved by off-the-shelf communication hardware, such as WiFi-enabled mobile devices. Such localization systems are more cost-effective. Patternmatching localization does not rely on any range estimation between mobile devices and infrastructure networks. For example, a system can be based on WiFi access points at unknown locations to serve as beacons. All the above works assume that beacons are reliable. In reality, some beacons may be moved to locations where they are not supposed to be without being noticed. Some beacon signals may be blocked by new obstacles deployed after the training phase, making their signal strengths untrustworthy. Some beacons may even conduct malicious attacks if they are compromised. A

Special Issue I

Page 61 of 65

ISSN 2229 5216

International Journal of Advances in Science and Technology, Vol. 3, No.3, 2011 malicious beacon is one which is tampered or compromised by an adversary and which can provide false distance or angle measurements. A malicious attack can be conducted individually or cooperatively. The major sources of unreliability come from unnoticed movement of some of these tiny beacons or unnoticed deployment of obstacles after the training phase, which may lower some beacons signal quality. However, signal quality from beacons can always be correctly measured, unless they are being interfered by noise.

III.

PROPOSED BEACON MOVEMENT DETECTION SCHEMES

To solve the BMD problem, we propose four detection schemes, namely LB, NB, SSB, and SSR schemes. These schemes differ in their local processing rules of beacons and the corresponding decision algorithms at the BMD Module. In the LB scheme, each beacon reports its observed signal strengths, which are used by the BMD Module to compute each beacons current location. The result is used to compare against its original location. In the NB scheme, each beacon locally decides if some neighboring beacons have moved into or out of their communication coverage range and reports its binary observations to the BMD Module. The SSB scheme is similar to the NB scheme, but the definition of movement is according to a threshold of signal strength change. In the SSR scheme, a beacon does not try to determine whether a neighboring beacon has been moved or not. Instead, each beacon reports the amount of signal strength change of each neighbor; the sum of all reported values is used by the BMD Module to make a global decision. A. Location Based Scheme The LB scheme assumes that the initial locations of beacons are known by the BMD Module in advance and utilizes localization techniques to monitor the locations of beacons. Each beacon is in charge of reporting the observed signal strength values of its neighbors to the BMD Module. The trilateration technique .Suppose a Beacon is moved out of its tolerable region. Since other beacons are unmoved, they can help to determine new location of the moved beacon. One point worth mentioning is that because of beacons movement, the estimated locations of other beacons may also be changed by a certain degree. So the outcome depends on the observations of the beacons. The LB scheme is sensitive to the performance of the adopted localization system. If the density of beacons is too low or signal strengths are too unstable, the results of movement detection cannot perform well. Since this scheme uses beacons to localize each other, moved beacons will also contribute some errors to the mutual localization process and thus influence our decisions. After the BMD Module receives the observations from all beacons, it estimates their possible locations under current mutual observations. Then the beacon with the longest moved distance will be selected. If the beacons current location is out of its tolerable region, it will be included in BMD Module and any observations contributed from Hi will be removed from boundary region. This process will be repeated until the suspicious beacons are found and are regarded as an unmoved one. B. Neighbor Based Scheme In the previous LB scheme, we report the observations according to the received signal strengths directly. It is sensitive to any slight movement. Hence, the NB scheme is designed to hide the information of signal strengths and just report binary observations to the BMD Module. In this scheme, each beacon Hi monitors the change of neighbor-hood relations with other beacons in its coverage area. The neighborhood relation of Hi at time t is defined as = An example with four beacons is shown in Fig.2 a, where the coverage of each beacon is a circle of

Fig 2. NB scheme: (a) the original relation, (b) a movement scenario, (c) observation matrix, (d) another movement scenario and Then the observation matrix is as shown in Fig. 2c. Then another movement of beacons is shown in Fig2.d in the NB scheme, our assumption that unreliable beacons are only a small proportion among all beacons. This assumption is reasonable because, in practice, beacons are usually moved by accident. Hence, we will try to construct a set BD that is as small as possible. First, we transform matrix observation graph as shown in Fig 2.e. After constructing graph, the NB scheme adopts a heuristic

Special Issue I

Page 62 of 65

ISSN 2229 5216

International Journal of Advances in Science and Technology, Vol. 3, No.3, 2011 approach as follows. If a beacon bis in-degree in graph is higher, it is more suspicious to be moved. So the Module sorts the vertices in graph according to their in-degrees of the uncovered edges in a descending order, and then selects the first one. This node is included in BD if any edge incident to it has not been covered. After selecting the most suspicious one, we will sort the vertices again. This process is repeated until a vertex cover is found. C. Signal Strength Binary Scheme In the previous NB scheme, we only consider the neighborhood relations between beacons. The LB scheme is more accurate because it considers the change of locations of beacons. In the SSB scheme, we assume that beacons can measure the signal strengths of packets from their neighbors. However, beacons do not report these measurements to the BMD Module directly. Instead, each beacon bi evaluates the amount of signal strength change of each neighboring beacon Hj locally and only reports a binary value to the BMD Module. Let the observed signal strength by Hi on Hj at time to be Si,j. The observation of Hi on Hj is = For each neighboring beacon Hi, we measure the average signal strength at each of these sampling points, assuming that Hj is moved to this sampling point. Note that if beacon bi does not hear any signals from Hj at a sampling point, we let its average signal strength be Smin. The major difference between the NB scheme and the SSB scheme is the calculation of local observation. However, the ambiguity property still holds D. Signal Strength Real Scheme Similarly to the previous SSB scheme, the SSR scheme assumes that beacons can measure the signal strengths from their neighboring beacons. However, in this scheme, the real signal strength variations, instead of binary values, observed by a beacon are reported to the BMD Module. The RSSI in both the Schemes cannot be measured accurately. Specifically, the observation is =| | E. Difference Correction Scheme Actual distances between Ho and H1,H2,.Hn are Do1,D02,.Don respectively difference measurement distance between object node O and the beacons H1,H2,Hn are D1,D2,..Dn respectively. The difference distance localization is measured from Difference correction and Maximum likelihood estimation method. IV. SIMULATION RESULTS

The sensing field is a 300 m X 300 m area. There are20 beacons randomly deployed on this field with the restriction that the distance between any two beacons is at least 5 m. This restriction is to avoid some beacons being placed too crowded, thus reducing the detection capability of the network. When a scenario violating the restriction is generated, we will discard it and regenerate another one.

Fig.3a Evaluation of hit and false probabilities for the SSB Scheme

Fig.3b Evaluation of hit and false probabilities for the SSR

Based on the received signal strengths at a distance of d is modeled by (d)= (d) +N (0, )

Special Issue I

Page 63 of 65

ISSN 2229 5216

International Journal of Advances in Science and Technology, Vol. 3, No.3, 2011 Where is the transmit power, which may vary among different hardware, is the path loss, which has a non isotropic and continuous property, and N (0, ) is a zero-mean normal random variable with a standard deviation to stand for dynamically shadowing noise. Received signal strength introduces the variance of sending power (VSP) to model the impacts of hardware difference and remaining battery of a device on transmit power. *(1+N (0, VSP) Where Pt denotes the initial transmit power and N (0, VSP) is a zero-mean normal random variable with a standard deviation VSP. The parameter VSP controls the degree of variance of sending power among different beacons. Each beacon randomly selects its when the simulation starts. The irregularity of signal fading is a common phenomenon. However, most path loss models do not take this non isotropic property of signal coverage into consideration. To capture this effect, RIM imports a degree of irregularity (DOI) to control the amount of path loss in different directions. (d)=PL (d)* Where is the optimal obstacle-free path loss formulation. PL(d)=PL( )+10log(d/ ) is to model the level of irregularity at degree i Where d0 is the reference distance the coefficient All results are from the average of 20 experiments. To reduce the influence of noise, signal strength is calculated from the average of 50 HELLO packets

Fig. 4. Comparison of hit and false probabilities by varying the standard deviation,

Fig. 5. Comparison of hit and false probabilities by varying the degree of irregularity DOI

Fig. 6. Comparison of hit and false probabilities by varying the varied sending power VSP of the RIM radio model

Special Issue I

Page 64 of 65

ISSN 2229 5216

International Journal of Advances in Science and Technology, Vol. 3, No.3, 2011

Fig 7 Error statistics of beacon nodes involved in V.

Localization

CONCLUSION

In this paper, we have proposed a Beacon Movement Detection in wireless sensor networks for improving localization accuracy. This paper describes a situation where some beacon sensors localization procedure are moved unexpectedly, called beacon movement events. We propose to allow beacons to monitor each other to identify such events. Four schemes are presented for the BMD problem. Moreover, it is clear that SSB and SSR Schemes have some errors which is improved by Difference distance localization. As a Future work can be done on to the system localization if we can relocate those moved beacons. VI.

REFERENCE

[1] S. Misra, G. Xue, and S. Bhardwaj, Secure and Robust Localization in a Wireless Ad Hoc Environment, IEEE Trans.Vehicular Technology, vol. 58, no. 3, pp. 1480-1489, Mar. 2008. [2] D. Moore, J. Leonard, D. Rus, and S. Teller, Robust Distributed Network Localization with Noisy Range Measurements, Proc.ACM Conf. Embedded Networked Sensor Systems (ACM SenSys),pp. 50-61, 2004. [3] E. Olson, J.J. Leonard, and S. Teller, Robust Range-Only Beacon Localization, IEEE J. Oceanic Eng., vol. 31, no. 4, pp. 949-958, Oct.2006.. [4] A. Kushki, K.N. Plataniotis, and A.N. Venetsanopoulos, Kernel-Based Positioning in Wireless Local Area Networks, IEEE Trans.Mobile Computing, vol. 6, no. 6, pp. 689-705, June 2007. [5] P. Bahl and V.N. Padmanabhan, RADAR: An In-Building RF Based User Location and Tracking System, Proc. IEEE INFOCOM,vol. 2, pp. 775-784, Mar. 2000. [6] J. Burrell, T. Brooke, and R. Beckwith, Vineyard Computing: Sensor Networks in Agricultural Production, IEEE Pervasive Computing, vol. 3, no. 1, pp. 38-45, Mar. 2004. [7] T.H. Cormen, C.E. Leiserson, R.L. Rivest, and C. Stein, Introduction to Algorithms. MIT Press/McGraw-Hill, 1990. AUTHORS PROFILE
S.Arun was born on 18th october 1979 Tiruvanamalai, India. He is working as Assistant Professor in Vel Tech High Engneering College Chennai and pursuing Ph.D. in the School of Electronics and Electrical Engineering, Singhania University, Rajasthan, India. He obtained M.E. degree in Communication systems 2004, Anna University and B.E. degree in Electronics and Communications Engineering from University of Madras, Chennai, India in the year 2001. . He has also published papers on image processing, wireless communication in National and International Conferences. His current research interests are Robotics, . S.Raghavendran, He is working as Assistant Professor in Vel Tech Dr.RR Dr.SR Technical University,Chennai and Finished M.E., (CSE), Sri Sivasubramaniam College of Engineering, Affiliated to Anna university in 2010 and Finished his B.E (CSE) in Saveetha Engineering college, Affiliated to Anna university in 2008 and He presented papers in several national and international conferences . R.Krishnamoorthy, Finished his B.E in Sree Sastha Institute of Science and Technology , Affiliated to Anna university in 2009, He presented papers in several national and international conferences

Special Issue I

65 of 65

ISSN 2229 5216