The analysis of localization algorithm of unscented particle filter based on RSS for linear wireless sensor networks

In order to solve the problem of mobile robot localization for linear wireless sensor networks due to the effect of various noises, this paper proposes the solution by unscented particle filter algorithm using received signal strength (RSS). The anchors are deployed on both sides of monitored linear region according to the specific rules. After the mobile robot carrying the sensor node sends data to anchors and receives data from them, the position of it can be calculated by the unscented particle filter algorithm. We discuss the localization accuracy of a complex velocity pattern that the robot adjusts direction according the circle path. We analyze the effect of different amount of anchor on the localization accuracy. We make the simulation to evaluate the localization accuracy of the algorithm at different parameter assignments compared with the other particle filter algorithm. The simulation results show that the unscented particle filter algorithm has more localization accuracy at various parameter assignments.