Study Of Self-localization Technology Based On Monte Carlo In Mobile Sensor Networks

Wireless Sensor Networks is made up of a large number of sensor nodes, each ofwhich has sensing, communicating, and computing abilities. Depending on whether thenodes move or not, it can be divided into static and mobile sensor networks. Comparedwith static sensor networks, mobile sensor networks is widely used in military affairs,biological medicine, environmental monitoring relying on its flexible self-organizing andstrong robustness features. However, the practical applications’prerequisite is to acquire theposition information of nodes, otherwise they would be meaningless. Therefore, researchon the localization technology of mobile sensor networks has important theoreticalsignificance and extensive application value.In this thesis, the existing localization algorithms are elaborated and classified,especially the series of algorithms based on Monte Carlo, including the theoretical analysisabout Monte Carlo method and the application principle and implementation process ofMonte Carlo in the localization of mobile sensor networks in detail. Simultaneously, wepoint out the main problems in existing localization algorithms.Then, regarding the problem of low sampling efficiency in existing localizationalgorithms, an improved algorithm is proposed by producing virtual beacon nodes using thecrossover and mutation operators in Genetic Algorithm. Meanwhile, samples are providedwith different weighting values according to the distance information. Simulation resultsshow that the proposed algorithm can significantly improve the network positioningaccuracy and efficiency especially in sparse mobile networks.Finally, for the problem that a large number of samples are required in the localizationalgorithms, an adaptive sampling algorithm is proposed by introducing Kullback-Leiblerdistance. Concurrently, samples are provided with different weighting values according tothe connectivity information. The experimental results show that the improved localizationalgorithm can reduce the number of sampling and the time consumption while ensuring thepositioning accuracy.