An Improved Monte Carlo Localization Algorithm for Mobile Wireless Sensor Networks

Currently localization algorithms for mobile sensor networks are mostly based on Sequential Monte Carlo method. However they appear either low sampling efficiency or demand high beacon density requirement issues to achieve high localization accuracy. Aiming to solve the problems, we proposed an improved algorithm called Genetic and Weighting Monte Carlo Localization (GWMCL) in which we apply the Genetic Algorithm into Sequential Monte Carlo, which indirectly increases the density of beacon nodes via producing virtual beacon nodes. Besides, we also consider the weight of different beacon nodes, which means that the weight of each beacon node is related to the distance between beacon node and unknown node. The simulation results illustrate that the proposed algorithm achieve better performance than Monte Carlo Localization algorithm, especially in the situation with low beacon density. Furthermore, it also exhibits high sampling efficiency and localization accuracy in sparse mobile networks.