Study Of Mobile Sensor Networks Localization Based On Monte Carlo Method

Wireless sensor network is deployed by a large number of sensor nodes with selforganization and multi-hop in the monitoring area,the sensor nodes observe the monitoring information required within perception,acquisition,transmission,and processing in the monitoring area in a collaborative manner.Wireless sensor network involves many technical fields,its localization technology is the most basic premise and it is also the base of a lot of key technologies.Because when the obtained data is related to the corresponding position information,many applications of wireless sensor networks are meaningful.The mobile sensor networks belong to the wireless sensor network have a more widely used for its flexibility,adaptability and the greater self-organization.But its location technology have a lot of difficulties because of the moving nodes.Thus achieving high-precision,highefficiency,low-power mobile sensor networks localization is an important part of this thesis.Firstly,the mobile sensor networks' localization technology was discussed,and the typically node localization algorithms of mobile sensor networks are classified as well as summary.Then the thesis describes the typical mobile sensor networks localization algorithm-the basic principle of sequential Monte Carlo method and the specific application methods and the thesis analyzes the advantages and disadvantages.Then,the thesis puts forward an improved Monte Carlo localization algorithm because of the sequential Monte Carlo algorithm's problems,it has optimization anchor box based on RSSI and adaptive sampling(RCAMCB).Based on RSSI ranging technique and centroid theory,the unknown nodes defined in a more precise anchor box and it has adaptive sampling by introducing the Kullback-Leibler distance and the area of sampling box in order to improve the efficiency of the algorithm.According to the influence of the anchor node,it optimizes the weight distribution.Then it calculates the location of the node.Finally,from a large number of MATLAB simulation,it verifies and analysis the positioning effect of the method.From the simulation results,it fully verifies the proposed algorithm can improve targeting efficiency,reduce energy consumption and can ensure high quality positioning accuracy.