An Energy Efficient Localization Strategy using Particle Swarm Optimization in Wireless Sensor Networks

Transcription

1 An Energy Efficient Localization Strategy using Particle Swarm Optimization in Wireless Sensor Networks Ms. Prerana Shrivastava *, Dr. S.B Pokle **, Dr.S.S.Dorle*** * Research Scholar, Electronics Department, G.H.Raisoni College of Engineering, Nagpur University, India ** Electronics & Communication Department, RCOEM, Nagpur University, India *** Electronics Department, GHRCE, Nagpur University, India Abstract The location information is very important in many wireless sensor network applications where in from which node the data is generated needs to be determined. Therefore it becomes necessary to know the location of each and every sensor that is deployed in the network. In our proposed Localization technique we have deployed static sensors in the geographical area of the network which are used for data gathering process and then tried to estimate the location of the mobile sink which is relocating itself at regular intervals in order to enhance the overall network performances. Some sensors with the known co-ordinates are deployed in the network and using their information the mobile sink will try to estimate its distance from these sensors and then from this compute its own location which is followed by applying a particle swarm optimization algorithm in order to reduce the error accumulation. The simulation of our proposed Localization strategy has been done in the NS-2.32 environment and its performance has been evaluated on the basis of various parameters of wireless sensor network. The simulation results showed a significant amount of improvement on the various performance metrics of the network. Keywords Localization Strategy, Particle Swarm Optimization (PSO), Received Signal Strength Strategy (RSSIS), Wireless Sensor Network (WSN). I. INTRODUCTION A Wireless Sensor Network consists of closely deployed sensors that are used for data gathering purpose. The primary task of any sensor is sensing, monitoring and relaying. The sensors are battery operated devices and hence have constraints like limited power, storage capacity and energy resources. Therefore it becomes necessary that all these limitations must be taken into account while designing any protocol in case of wireless sensor network. [1], [2]. The applications like tracking and routing in case of wireless sensor networks are completely dependent on the position of the various nodes that are deployed in the network. So the localization technique plays a very vital role for such applications and the accurateness of such techniques is extremely important [4]. Any basic localization algorithm would estimate the location of the nodes based on the input information that is provided by the neighboring nodes in order to estimate the location or position of the unknown node. Based on this information it will employ various ranging techniques in order to estimate the distance and from this it will compute the position of the unknown node [5]. There are two types of Localization Techniques that are widely used. The first is the range based localization techniques and the second is the range free localization techniques. The time of arrival, the time difference of arrival, the angle of arrival and the received signal strength are the various range based localization techniques which map the above mentioned parameters between the sender and the receiver in order to estimate the distance [6].In the second type of Localization technique which is the range free localization technique usually the position is estimated with reference to 17

2 the other neighboring nodes that are deployed in the network. Using their location information the location of the unknown node can be estimated. For this an efficient hop by hop algorithm has been proposed which works on this concept [7]. II. RELATED WORK D.Niculescu, B.Nath [8] proposed the basic fundamental range based localization system in GPS which employs a satellite in conjunction with the ground station as a reference node in order to provide the positioning services for the nodes that are deployed in the network. But because of the large volume of the nodes attaching a GPS receiver to each and every node in WSN was quite expensive. Moreover it was resulting in the large amount of positioning errors. Hence there was an urgent demand for efficient and cost effective localization algorithms in WSN S. Po Jen, Cheng Pei Wu [9] discussed the range based localization method where the location of the node is computed with respect to the other sensors in its surrounding area. They employed RSSI technique which is used to predict the distance between the unknown node and its neighbor according to the received signal strength and path loss pattern. After this they employed PSO in order to eliminate the error accumulation. But the weak point of RSS was its sensitivity to uncertain environmental elements like obstacles, noises which may produce large errors. K.Liu,X.Yan,F.H [10] proposed a distributed hop by hop positioning algorithm,i.e,dv Hop where the sensors will broadcast their location information throughout the network and each sensor will keep track of hop count to each of these sensors. After this the average distance is calculated and broadcasted. After this two step broadcasting each node can estimate its distance to each sensor with known co-ordinates, i.e., the product of average hop distance and corresponding hop counts in order to estimate the location. Although DV hop is simple and useful it can provide good position estimation if the node distribution in the network is dense and uniform. Also DV Hop introduces lots of traffic load and communication delay in WSN due to the fact that it works in two steps. III. PROBLEM IDENTIFICATION The limited power supply to the nodes is the main concern for designing any protocol in case of wireless sensor networks. The localization strategy should be such that it should not affect the energy efficiency of the network as well as provide with the accurate result. The range based localization techniques like time of arrival, time difference of arrival and angle of arrival are suitable for short range communication. On the other hand the received signal strength indicator method is highly sensitive to the environmental conditions which may introduce large amount of errors. The other range free localization techniques were introducing a substantial amount of delay. Moreover if there are insufficient amount of anchor nodes deployed in the network then localization becomes difficult and inestimable thereby resulting in reduction in the localization accuracy. Also the iterative process that is involved during the search process result in the error accumulation which grows around the neighborhood of the sensor. Thus the main focus is to devise a localization strategy which would result in the reduction in the hardware cost, improve the localization accuracy and at the same time it should minimize the error accumulation which is resulting from the iterative process. 18

3 IV. PROPOSED METHODOLOGY In other to overcome the above mentioned drawbacks we are proposing an energy efficient Localization Strategy using particle swarm optimization in wireless sensor networks. In our system model we have deployed various sensors in the geographical extent of the entire network which are used for the data gathering purpose. The sensors will organize themselves into various clusters depending upon their location and one mobile sink is deployed in each cluster. The mobile sink will relocate itself at regular interval of time in order to collect the data from the various sensors. This would results in the significant amount of energy saving thereby increasing the life of the sensor and hence minimizing the traffic intensity. In our model the sensors are static and the sink is mobile. The position of sink is changing at regular interval of time. The location of the mobile sink needs to be estimated and needs to be broadcasted to the sensors as well as the other mobile sinks in different clusters from time to time. In each and every cluster we have deployed three sensors with known position. Let (X1, Y1), (X2, Y2) and (X3, Y3) be their co-ordinates and (X, Y) be the coordinates of the mobile sink. Then the distance between this sensors and the mobile sink is estimated as : D1 Y ( ( X1 X ) ( Y1 ) ) (1) D2 Y ( ( X 2 X ) ( Y2 ) ) (2) D3 Y ( ( X3 X ) ( Y3 ) ) (3) The above non linear equations are then solved in order to estimate the co-ordinates of the mobile sink, i.e., (X, Y).Then on this known and recorded position the particle swarm optimization algorithm is applied in order to obtain the best location. The PSO algorithm is employed in the localization process of our technique in order to obtain the optimized location of the unknown sensor.in PSO two process that is the searching and the optimizing process are involved. Our Localization scheme adopts the weighted PSO. Let M be the unknown mobile sink. It is getting the information from the sensors with the known coordinates S1, S2 and S3.The inexact distance between the sensors and the mobile sink be D1,D2 and D3 respectively. The difference between the real and estimated location of the unknown mobile sink is given by the error equation. The mobile sink M will generate K particles with K random co-ordinates. Each of the K particles will calculate the error value and will record as particles best value for each random value. Let this value be denoted by Pbest.This value is updated during the search process and is described as the fitness value for the mobile sink. The smallest fitness value is recorded as the group s fitness value which is indicated by Gbest.Once Pbest and Gbest are obtained the particle will calculate its next position by then recalculate the fitness value and then update these new values. This iterative process will repeat until the convergence is attained. The pseudo code for our localization algorithm is as stated below: { Listen and collect known sensor position information if three sensors with known position discovered 19

4 then call procedure localization Procedure Localization { Estimate the distance and the location Employ PSO Update information Broadcasts the location information With transmission radius and updated position Localization complete exit } } graph.the performance of our proposed LSPSO has been done with the existing technique, i.e, ERSSM which is based on the received signal strength. The average energy consumed, end to end delay and the packet delivery ratio are the various performance metrics that are considered for evaluating the performance of the proposed method. The X graph were observed for the average energy consumption, end to end delay and the packet delivery ratio by varying the number of the sensors as shown in the figure below. VI. SYSTEM MODEL AND PARAMETERS The proposed Localization strategy using particle swarm optimization (LSPSO) has been evaluated in the NS 2.34 environment. The sensors are deployed in the bounded region of 1000 x 1000 sq.m. using uniform distribution. The DCF is used as the fundamental MAC technique for the wireless LAN.The channel capacity for the mobile host is set to 3Mbps.The transmission and sensing range of the sensors is set to 150m.In the simulation the routing protocol used is adhoc on demand distance vector routing protocol and the traffic generator employed is CBR. Fig.1 : Energy Consumption Vs.Number of Sensors VII. SIMULATION RESULTS The TCL script for our proposed algorithm is run under NS-2.32 environment. The NAM output will give us the visualization of our network model. For tracing and monitoring our simulation we will run our trace file which we have set in the TCL script. The trace values obtained will be analyzed by making use of the trace data analyzer that is the X- Fig.2 : Delay Vs.Number of Sensors 20

5 Fig.2 : Delivery Ratio Vs.Number of Sensors From the above figures 1, 2 and 3 we observe that the proposed LSPSO shows an improvement in the packet delivery ratio, minimizes the end to end delay and achieves a significant amount of reduction in the average energy consumption as compared to the existing ERSSM.The observations were made by varying the number of sensors. VIII. CONCLUSION In this paper the proposed energy efficient localization strategy for wireless sensor networks minimizes the implementation cost. The mobile sink will estimate its distance with respect to the neighboring nodes and then from the estimated distance compute its own location. In order to minimize the location inaccuracy the particle swarm optimization is applied to the estimated location which would improve the accurateness of the proposed strategy. From simulation results it is observed that the proposed localization technique improves the various performance metrics of the wireless sensor network in comparison to the existing received signal strength method. REFERENCES [1] Mohamed K.Watfa, Sesh Commuri, An Energy Efficient Approach to Dynamic Coverage in Wireless Sensor Networks, Journal of Networks,vol.1,No.4,August [2] Khin Thanda Soe, Increasing Lifetime of Target Tracking Wireless Sensor Networks, World Academy of science, Engineering and Technology, vol.44, August [3] Amitangshu Pal, Localization Algorithms in Wireless Sensor Networks: Current Approaches and Future Challenges, Network Protocols and Algorithms, vol.2, No.1, [4] Vibha Yadav, Manas Kumar Mishra, A.K. Singh, M.M. Gore, Localization Scheme for three dimensional wireless sensor networks using GPS Enabled Mobile Sensor Nodes, International Journal of Next Generation,vol.1,No.1,December [5] Ewa Niewiadomska Szynkiewicz,Michal Marks,Mariusz Kamola, Localization in wireless sensor networks using heuristic Optimization Techniques, Journal of Telecommunications and Information Technology,pp ,2011. [6] Cesare Alippi,Giovanni Vanini,Politecnico di Milano, A RSSI based and calibrated centralized localization technique for wireless sensor networks, proceedings of IEEE International Conference on Pervasive Computing and Communications, March [7] Engin Mas Azade,Ruixin Niu,Pramod K.Varshney, Mehmet Keskinoz, Energy Aware Iterative source Localization for wireless sensor networks, IEEE Transaction on Signal Processing, vol.58, No.9, September [8] D.Niculescu,B.Nath, Adhoc Positionng Systems(APS),proceedings of IEEE Global Telecom conference,globecom 01,Nov [9] Po Jen Chuang, Cheng Pei Wu, Employing PSO to enhance RSS Range Based Node Localization for wireless sensor networks, Journal of Information Science and Engineering,pp ,December

6 [10] K.Liu,X.Yan,F.Hu, A Modified DV Hop Localization Algorithm for Wireless Sensor Network, proceedings of IEEE International conference on Intelligent Computing and Intelligent Systems,pp ,2009. [11] Hongyang Chen, Qingjiang Shi, Rui Tan,H. Vincent Poot, Kaoru Sezaki, Mobile Element Assisted Co-operative Localization for wireless sensor networks with obstacles,ieee Transaction on wireless communications, vol.9,no.3,march [12] Yug Wang, Xiaodong Wang, Demin Wang, Dharma P.Agrawal, Range free localization using expected Hop progress in wireless sensor networks, IEEE Transaction on Parallel and distributed Systems, vol.20, No.10, October 2009 [13] Network Simulator: AUTHOR S PROFILE Dr. Sanjay. B. Pokle obtained his Bachelor s degree in Electronics and Telecommunication Engineering from Govt. College of Engineering, Pune University, India in 1993.He then obtained his Masters degree in Electronics Engineering and also Ph. D. in Electronics from Visvesvaraya National Institute of Technology Nagpur, India. His research area includes designing aspects of MIMO-OFDM Wireless Communication Systems and Wireless channel Estimation Algorithms. He has published research papers in the reputed national and international Journals and also has presented papers in the reputed national and international conferences. He has guided several projects in the area of signal processing, Digital image processing, Artificial intelligence etc at post graduation level. He is member of technical societies like ISTE and IEEE. He has total 18 years of experience which includes 3 years industry and 15 years of teaching experience. Presently he is working as Professor and Head of Electronics & Communication Engineering Department, Shri Ramdeobaba College of Engineering and Management, Nagpur. Mrs. Prerana Shrivastava obtained her Bachelor s degree in Electronics Engineering from University of Nagpur, India. Then she obtained her Master s degree in Electronics Engineering and currently pursuing her PhD in Wireless Sensor Networks from G.H.Raisoni College of Engineering, Nagpur, India.She is working as an Assistant Professor in Electronics Department at Lokmanya Tilak College of Engineering, University of Mumbai, India. Her specializations include Electromagnetic Wave Theory, Image Processing and Computer Networks. Her current research interests are Wireless Adhoc Networks, Wireless Sensor Networks and Network Security. 22