Reserch On The Key Technology Of WIFI Indoor Localization

In recent years, with the development of wireless communication technology, the positioning service in the people’s production and life plays a more and more important role. With the maturity and wide application of the outdoor positioning technology, the demand for indoor positioning is increasing day by day. But indoor positioning technology is not mature, there is a need to overcome technical difficul ties.In this thesis we consider the problem of indoor positioning technology based on location fingerprint. Firstly, through the experimental data, the characteristics of the space, probability distribution and convergence rate of the received signal stre ngth are analyzed. The classic indoor positioning technology is introduced, and the characteristics of each technology are analyzed. The problems existing in the existing positioning system are analyzed. For example, the indoor positioning system with high accuracy need more APs, the positioning algorithm does not make full use of the information of location fingerprint database. In view of these problems, some improvements are put forward. According to the local information of the location fingerprint, the concept of local degree of dispersion is defined, and the distance measure criterion of the signal strength vector is proposed. Associating the information of signal space and location coordinate, the selection criteria of location fingerprint is proposed to improve the positioning performance.Analyzing clustering algorithms used in indoor positioning in recent years, the feasibility and validity of clustering according to the spatial coordinates are explored. Based on static indoor positioning algorithm, limiting the region around current location with historical location information, we propose a dynamic regional prediction algorithm. Simulation and analysis show that the proposed algorithm is effective and robust. Associating coordinate with signal space, a dynamic indoor positioning system is proposed by combining the proposed localization algorithms.