Using BP Neural Network Algorithm Improve the Prediction Accuracy of Distance between ZigBee

Abstract

This paper proposes that we can improve the accuracy of the distance between ZigBee with the algorithm of BP neural network. Firstly, we analyze the wireless signaling path loss model, the principle of measuring distance based on RSSI, and the algorithm of BP neural network fully. Secondly, we get the experimental data from the hardware platform of ZigBee. Finally, we use the algorithm of BP neural network to emulate and analyze the experimental data. Based on the results of the emulation and analysis, the maximum error of the distance prediction within 120m is 7.62%, with the prediction accuracy improved significantly.