采用混合指纹处理方法的定位系统性能分析

In the field of Fingerprinting Location (LF), the quality of the fingerprint has a great influence on the location results. In this paper, we use a variety of fingerprint processing methods to improve the positioning accuracy. In the offline stage, Fingerprint merging and fingerprint filling are used to ensure the uniform distribution of the fingerprint in the space, and the online stage uses the heuristic fingerprint feature dimension rules to select the candidate fingerprints, then reduce the location area, and last determine the weight of the candidate samples by fuzzy clustering. In the end, we test the performance based on a real cellular network in a dense urban area, the results show that this method is effective in the actual network environment.