Weighted positioning method based on RSSI probability distribution and Bayesian estimation

Aiming at the problem that the traditional Wi-Fi positioning technology based on distance measurement does not consider the distribution characteristics of received signal strength indication (RSSI) values, which may result in poor indoor positioning results, this paper proposes a weighted positioning method based on RSSI probability distribution and Bayesian estimation. On the basis of studying the stationarity and distribution characteristics of RSSI. The method introduces the prior RSSI probability distribution into to the calculation of weight through Bayesian estimation. It can also give lower weights to outliers go as to reduce the impact of environmental noise and external uncertain factors on the positioning accuracy, then the position with the largest weight will be taken as the positioning result. Experimental results show that compared with results of trilateral localization, weighted centroid localization and weight correction algorithm, the average error of this method is reduced by 45.4%, 14.6%, 8.2%, and the error of cumulative probability within 50% is reduced by 66.7%, 42.1%, 32.4%.