Research Of Node Localization In Wireless Sensor Networks Based On Bat Algorithm

With the development of technology,wireless sensor network(WSN)as a new network technology integrated with sensor technology,modern wireless communication technology,distributed information processing technology,has been received a wide range of attention.It can be widely applied in healthy caring,military surveillance,intelligent transportation,environment monitoring.Among these application scenarios,the accuracy of node location information is a standard to measure the performance of the whole wireless sensor network.So the node localization is the key technology of wireless sensor network.In order to ensure the accuracy of localization in WSN the node localization algorithm is usually used based on signal strength indicator(RSSI),time of arrival(TOA),time difference of arrival(TDOA)and angle-of-arrival(AOA).After getting the distance between the nodes,the location coordinates of the unknown nodes are calculated by the methods such as three edge positioning,triangle localization and maximum likelihood estimation.However,in these positioning algorithms,the location error which cannot be eliminated completely has greatly affected on the localization accuracy.Then,the bat algorithm and its improved algorithms are introduced to the location optimization of wireless sensor nodes so that the localization accuracy of the unknown nodes could be improved obviously by offsetting the range error.The main research work of this paper is as follows:(1)Researching on the global convergence of the bat algorithm and its application to wireless sensor network node localization.Based on the deeply research of the theory of node localization in WSN,this paper discusses the optimization problem using the bat algorithm by transforming the node location problem into a mathematical optimization problem.The evolution of bat algorithm could be considered as a random process,which meets the condition of Markov process.The Markov chain model is introduced into the basic bat algorithm.The basic mathematical definition about the bat algorithm is also given in this paper and Markov chain model is used to demonstrate the reducibility and homogeneity of bat population state space and analyze the convergence of bat algorithm theoretically.The result shows that bat population sequence could converge to the optimal solution in probability 1.Subsequently,the bat algorithm is applied to the WSN node localization.The experimental results show that the algorithm is better than other algorithms,and the positioning accuracy of the algorithm is also improved.(2)A new node localization method based on the self-learning ability of the bat algorithm(SV-BA)is proposed.Because the basic bat algorithm is easy to be trapped in local optimum and has the slower convergence rate in later procedure,a bat algorithm with the capability of self-learning and individual variation is proposed.In this proposed algorithm,the best global individual with the self-learning capability can self-optimize within a small range of solutions and could lead to other individual develop deep searching.In addition,the each individual will generate a dynamic number variation cluster in proportion its fitness value.According to the rule of greedy selection,the best individual in the variation cluster was selected which protects the excellent individual,increases the population diversity and avoids the individual degradation.The proposed algorithm makes use of the self-learning and individual variation improved the optimization accuracy and convergence speed.The simulation results for the standard test functions show that the improved bat algorithm has significant advantage of high optimization ability and searching precision,and can skip from local optimum effectively.Then taking the improved bat algorithm to solve the node localization problem in wireless sensor networks,the experiment indicate that the positioning accuracy of the nodes is further improved,and has great value in engineering of complex function optimization.(3)A new WSN node localization method based on multi-agent bat algorithm(MA-BA)is proposed.In order to improve the further performance of the bat algorithm,the new algorithm based on multi-agent is proposed which integrates the multi-agent technology,and makes full use of the individual initiative and individual interaction,so that the global search ability of the algorithm is improved.In the proposed algorithm,the every bat individual is an agent,which could compete and cooperate with its agent neighbor areas to improve the efficiency of local searching.The algorithm can avoid local optimization and accelerate the convergence rate of the algorithm.Simulation results for standard test functions indicate that the proposed algorithm remarkably improves the global optimizing ability and evolutionary efficiency compared to other algorithms.Through implementing the MA-BA to node location prediction,the precision of the unknown node location could be improved due to decreasing the ranging error and has certain significance to practical application of wireless sensor network node localization.