Optimized Radial Basis Function Network Based on Beetle Antenna Search for Indoor Localization Algorithm

Abstract

In order to improve the accuracy and robustness of Bluetooth indoor localization, an improved fusion hybrid filter and radial basis function neural network (RBFNN) indoor localization method is proposed, which effectively improves the accuracy of received signal strength (RSS) by combining various filtering algorithms, and introduces a radial basis neural network in machine learning algorithm to map the nonlinear relationship between RSS and anchor node to signal receiver localization. RBFNN is optimized using the algorithm of beetle antenna search to further improve the stability of localization. An experimental platform based on NRF52810 and a smart phone is built to verify the proposed method. Theoretical analysis and experimental results show that the average positioning error of the proposed method is 0. 63m, the confidence probability of less than lm is 75%, and the confidence probability of less than 2m is 96%. It can effectively reduce the positioning error and improve the positioning accuracy, and is easy deploy, which has high application value.