Varying the Sample Number for Monte Carlo Localization in Mobile Sensor Networks

Monte Carlo method has been widely used in many fields in the past few years. Currently, for mobile object localizing and tracking, Monte Carlo method has been practically proved a successful solution to solve these non-Gaussian, non-nonlinear and multi-dimensional systems. Recently, several Monte Carlo localization algorithms have been proposed for mobile sensor networks which point out a new direction for localization in sensor networks. However, these previous literatures generally use a fixed sample number in their Monte Carlo localization algorithms which is very inefficient and inappropriate to the low energy low computational capability sensors. In this paper, we introduce a sample adaptive Monte Carlo Localization algorithm (SAMCL) to improve the localization efficiency. Simulation results demonstrate that our method produces good localization accuracy as well as low computational cost compared with the previous Monte Carlo localization algorithms.