The Research Of Indoor Positioning Algorithm Based On WIFI Networks

With advances in technology and the development of society,Location Based Services has gradually extended from outdoor scene to indoor environment. As a result, it has a broad prospect. A good positioning technology is a necessary premise on its application and development. With the rapid development of wireless LANs, WLAN-based positioning technology is becoming mainstream indoor positioning technology, which Wi Fi indoor positioning technology that based on the location fingerprint is simple, low cost, high positioning accuracy, robustness, scalability, etc, has become a hot spot of indoor positioning technology. This article has analyzed the development status of indoor positioning technology based on WLAN and conducted in-depth research for localization algorithm based on Location fingerprint in the Wi Fi network.Firstly, this article has studied the location fingerprint positioning technologies commonly used in neighboring positioning algorithm in detail. The core of neighbor localization algorithm is to solve the Euclidean distance. In the past, the result of Euclidean distance does not consider that the signal strength of each AP has different importance of positioning results. Euclidean distance is not a true reflection of the similarity between the two fingerprints. To solve the above problems, this article presents the neighbor localization algorithm based on weighted Euclidean distance. This algorithm has given different weights to the components of the signal strength of each AP in Euclidean distance to make it be more in line with the actual situation.Secondly, this article also has analyzed the statistical of Wi Fi signal strength. The signal strength can be approximated as a lognormal distribution and then we have preprocessed signal strength. According to this basis, the article proposes a positioning algorithm based on the area of overlap of the lognormal distribution of fingerprint localization algorithm. The principle of the algorithm is to calculate the area coincidence degree of probability density function between end-user fingerprint and database reference fingerprint. The coincidence of the size of the area will be as a measure of the similarity of fingerprints.Finally, we can use the signal strength data collected in the actual environment to validation the both indoor positioning algorithms proposed in this article on MATLAB platform. Then compared with the positioning results of traditional fingerprint localization algorithm, the positioning results verify the effectiveness of the algorithms.