最强基站MAC地址匹配的RSSI加权室内定位方法

Indoor positioning method based on strongest base station MAC address matching and RSSI weighting

针对传统k近邻算法定位时不能有效剔除距离较远参考点的问题,提出最强基站介质访问控制(MAC)地址匹配的接收信号强度指示(RSSI)加权改进室内定位方法:离线阶段,通过模糊c均值算法划分待测点的定位区域,生成基于区域划分的聚类指纹库;在线阶段,首先确定待测点所在的目标区域,其次在目标区域内利用动态加权k近邻算法剔除距离偏远的参考点,然后通过MAC地址序列匹配的方法,只信任最强的基站,进一步筛选出k个中最优的参考点,最后计算最优参考点对应坐标的加权平均值作为待测点的最终估计位置。实验结果表明,与动态加权k近邻算法相比,该算法在房间以及走廊环境下的平均定位误差都有改善,并且1~2 m和2~3 m定位精度的可信度有较好的提升。

Aiming at the problem that the traditional k-nearest neighbor algorithm can not effectively remove the reference point with a long distance in positioning,the paper proposed an improved indoor positioning method based on the strongest base station media access control(MAC)address matching and received signal strength indicator(RSSI)weighting:in the offline phase,the fuzzy c-means algorithm was used to divide the location of the point to be measured,and a clustering fingerprint database based on the region division was generated;in the online phase,the target area where the point to be measured was located was determined firstly,and in the target area,the remote weighted reference point was eliminated by using the dynamic weighted k-nearest neighbor algorithm secondly,then,through the method of MAC address sequence matching,only the strongest base station was trusted,the optimal reference points among k were further selected,finally,the weighted average of the coordinates of the optimal reference point was calculated as the final estimated position of the point to be measured.Experimental results showed that,compared with the dynamic weighted k-nearest neighbor algorithm,the average positioning error of the proposed algorithm in the room and corridor environment would be both improved,and the reliability of the positioning accuracy in 1-2 m and 2-3 m could be significantly promoted.