Research Of Bayesian Indoor Positioning Algorithm Based On ZigBee

Wireless positioning technology based on ZigBee’s RSSI is one of the maintechniques of indoor positioning system. Because of the complexity and variabilityof indoor environment, positioning algorithm using RSSI maybe causes percent50ranging errors. Therefore, we improve the algorithm in view of the coarsepositioning mode.The paper analyses two main factors, ranging error and error of calculation,which mainly influence positioning accuracy. First, it assays the indoor wirelesschannel transmission loss model which influences the ranging error of RSSI. Second,it analyzes the triangle centroid algorithm, and the weighted triangle centroidalgorithm, which lead to the error of calculation. Third, according to the feature ofrandom and normal distribution of RSSI, Gauss filtering algorithm and Bayesianstatistics theory are analyzed in order to improve RSSI-based positioning algorithm.The proposed algorithm has three steps. First, RSSI values between theunknown node and anchor’s are read from Zigbee wireless network which iscomposed by the wireless chip CC2430. Then, the read RSSI values are processed byGauss filtering algorithm. Distance between nodes is calculated through theempirical model. Initial uncalculated position information is estimated by thepositioning algorithm. Finally, establishing a Bayesian filtering model from theprinciple of the probability statistics, it takes the estimated location information asthe prior value, to further improve the indoor positioning accuracy. The experimentalresults show that the positioning accuracy has been improved after improvementalgorithm. Positioning accuracy can be accurate to about0.01~1meter.