Research on Node Location Algorithm of Zigbee Based on PSO-GRNN Neural Network

Abstract

The parameters of the traditional wireless signal propagation model are generally obtained by fitting or directly based on experience, and are affected by factors such as complex environment and multipath effect, which makes the positioning accuracy is not high. In view of this, a new node location algorithm of Particle Swarm Optimization-Generalized Regression Neural Network (PSO-GRNN) was presented. The RSSI received at each reference point is taken as the input of the network, the position coordinates of which are used as the output of the network to construct the GRNN, training the neural network by PSO algorithm with linear decreasing inertia weight, and the best Spread is cycled, avoid interference from human factors when adjusting this parameter, finally the trained model using the forecast point positioning. Simulation verified by MATLAB and Zigbee experiments, compared with the non-optimized GRNN model and BP neural network model, this algorithm has higher positioning accuracy in node localization.