Research on improvement to WiFi fingerprint location algorithm

For improving the precision of the fingerprint positioning algorithm based on WIFI and decreasing the complexity of the algorithm at the same time, this paper proposed a KNN + Bayes fusion positioning algorithm based on subsection and interpolation. In the process of building fingerprint library, we use the method of linear interpolation to get the fingerprint which is in the middle of adjacent sampling points. We improve the intensity and positioning precision accompany with reduced workload of training data acquisition. In the process of fingerprint matching, we do MAC filter first to narrow the scope of fingerprint matching. For improving the precision of the fingerprint positioning algorithm based on WIFI and decreasing the complexity of the algorithm at the same time, this paper proposed a KNN + Bayes fusion positioning algorithm based on subsection and interpolation. In the process of building fingerprint library, we use the method of linear interpolation to get the fingerprint in the middle of adjacent sampling points; we improve the intensity and positioning precision with reduced workload of training data acquisition. In the process of fingerprint matching, we do MAC filter first to narrow the scope of fingerprint matching. And then we use KNN algorithm to the rough positioning in the scope to get the credible region of the user. We also use interpolation method to generate virtual sampling fingerprint to increase the density of fingerprint. At last, we use Bayes algorithm to estimate the accurate location of target. We do positioning experiment in real scenario, using 3 methods for 30 groups positioning experiment separately and selecting 10 points randomly for error analysis, the results show that this algorithm decreases time complexity O(4n/5) compared to Bayes algorithm and increase the positioning precision by about 6% than KNN algorithm.