Abstract:

 Positioning is the basic function of Wireless Sensor Network.The paper presents an algorithm that optimizes the distance data measured in ranging stage by combining the PSO algorithm and BP neural network.By taking the PSO algorithm as a learning algorithm of BP neural network,it shortens the training time of BP network and speeds up the convergence rate of the algorithm.The simulation results show that the positioning accuracy is significantly improved by up to 27.3% over those by other algorithms.

Key words: WSN;BP neural network;particle swarm optimization;RSSI;localization algorithm