基于卡尔曼平滑的AWKNN室内定位方法

基于接收信号强度指示的WIFI室内定位方案存在采集信息跳变现象，进而影响定位精度的问题，提出一种基于卡尔曼滤波的改进自适应加权K最近邻（AWKNN）定位方法。对比分析多种平滑RSSI算法可行性，验证基于卡尔曼滤波对RSSI值进行平滑处理的优势，结合AWKNN算法并采用均方差计算匹配度，通过实时监控相匹配的无线接入点个数后自动调整均方差分母大小，以此实现定位误差的有效控制。实验结果表明，该基于卡尔曼的AWKNN算法在稳定性和定位精度方面较传统WIFI指纹算法有较大幅度提高。

AWKNN Indoor Location Methods Based on Kalman Smoothing

Aiming at the problem of the information jump in WIFI indoor location based on the received signal strength indication (RSSI), which influences the positioning accuracy, an improved adaptive weighted K nearest neighbor (AWKNN) localization method based on Kalman filter is proposed. In this paper, the feasibility of smoothing the RSSI algorithm is compared and analyzed, and the advantages of smoothing the RSSI based on the Kalman filter are verified. Combining with the AWKNN algorithm and taking advantage of the mean square deviation to calculate the matching degree, the size of denominator m in the mean square error can be adjusted automatically through monitoring the number of matching wireless access points in real time to achieve the effective control of positioning error. The experimental results show that the AWKNN algorithm based on Kalman filter is more effective than the traditional WIFI fingerprint algorithm in terms of stability and positioning accuracy.