An Indoor Positioning Method Based on RSSI Probability Distribution

In view of the influence of the time-variation of RSSI on the positioning accuracy in Wi-Fi indoor positioning, this paper proposes to use the probability distribution of RSSI value as a fingerprint feature over a period of time, and combines the dimension reduction algorithm and the weighted K nearest neighbor algorithm to achieve positioning. The method firstly calculates the probability distribution of the received RSSI value, uses the dimensionality reduction algorithm to reduce the dimension of the statistical probability distribution.The K-reference points with the smallest Euclidean distance were combined with the weighted nearest neighbor algorithm to obtain the positioning results. Through simulation experiments, it is shown that the positioning accuracy is higher than the traditional method, and the positioning time is significantly reduced.