Research On The Algorithms Of Localization And Coverage Holes Maintenance For Wireless Sensor Network In Complex Environments

Wireless Sensor Network(WSN)is a new type of information acquiring system,and consists of a large number of nodes which are deployed in the monitoring field.It is characterized as low cost,random deployed and distributed computing.It has been widely applied in military,civil,industrial,and other fields.Therefore,WSN has attracted more and more researchers' attention.In the complex environment,there are the features and difficulties such as non-line of sight(NLOS)propagation,moving node,multiple targets and damaged nodes occurred in the WSN.In order to solve these problems,the key technologies of node localization,target localization and network maintenance are investigated in depth in this dissertation.The main research content and results are presented as follows:Because the traditional GPS is not suitable to work in the indoor environment and the RSS messurment is vulnerable to environmental interference,the Expectation Maximization(EM)based indoor localization method is proposed.The value of Received Signal Strength(RSS)measurement between the beacon nodes is obtained,and the EM algorithm is employed to estimate the parameters of propagation model.Moreover the improved Tabu search method is used to estimate the position of unknown node and the experiment results show that the proposed method owns higher localization accuracy.Because Bluetooth technology ensures stable transmission in the noisy environment and is pre-installed in majority mobile devices,the indoor localization system is developed in Bluetooth.It provides an experimental platform for the proposed method.The non-line of sight(NLOS)propagation will occur when the communication channel is blocked by obstacles in indoor or some special circumstances.And the NLOS error has great effect on the measurement.Two node localization methods are proposed to identify NLOS propagation.Under the Time of Arrival(TOA)measurement,a NLOS localization method based on Pearsonchi-square test is proposed which identifies the propagation condition with the statistical analysis technique.Furthermore,the localization cost function is established and the position of unknown node is estimatedwith the particle swarm optimization algorithm.Under the TO A and RSS measurement,the NLOS error mitigation method is proposed based on the hypothesis testing.This method integrates the hypothesis testing method into Kalman filter.And the fusion algorithm is used to mitigate the effect of the NLOS error on the localization accuracy.The experiment results show that the two proposed methods can reduce the impact of NLOS error on the localization accuracy effectively.It is difficult to estimate the NLOS error and localization accuracy of mobile node is relatively low.Therefore the unknown NLOS error based mobile node localization method is proposed.Firstly,the performance of filtering algorithm is analyzed.The generalized likelihood ratio test method is proposed to detect the signal propagation condition.And the VB-AKF algorithm is applied to estimate the mean and variance of measurement noise in NLOS environment.Finally,the maximum likelihood estimation method is proposed to estimate the position of mobile node.Simulation results show that the proposed method will improve the localization accuracy in LOS/NLOS mixed environment.The traditional multi-target localization methods have relatively high computation complexity and require veryexpensive hardware configuration.In order to fix these problems,the low weight multi-source localization based on quantified measurement is proposed.The acoustic source propagation model is firstly analyzed.Then the quantification method is proposed.The maximum likelihood estimation method is used to establish the localization objective function.And the particle swarm optimization is employed to estimate the position of source.Finally the Kalman filter is assigned to predict the source position in the next time.Simulation results show that the proposed method can reduce the data transmission and estimate the source position with high accuracy.If the network is running for a long time,the coverage holes will probably occur.The coverage holes identification and repair methods based on Fuzzy C-mean are proposed.The node detection model based on Neyman-Pearson criterion is set up.And the spatial constraints fuzzy C-mean method is proposed to identify the coverage holes and it also estimates the area of coverage holes.Finally the coverage holes repair method based on Tabu search is proposed.Simulation results show that the proposed method can successfully repair the coverage hole and achieve the full coverage in the monitoring field effectively.The theories and methods of localization and coverage hole maintenance technologies have been systematically researched for wireless sensor network in this dissertation.In comparison with other correlative methods through the simulation experiments,the proposed methods have been verified to be feasible,available and advanced.