Research On The Indoor Tracking Based On Particle Filter Algorithm

As location-aware services on demand, high-precision indoor positioning becomes the core of location-based services. While the flexibility, mobility and simple point to point technologies, WLAN positioning system has gradually become a research hotspot. And in fact, real-time location information of dynamic users is also significant, based on it indoor tracking technology developed very rapid. Particle filter algorithm has shown its good performance in non-linear and non-Gaussian models and is paid more and more attention. Therefore, this paper studies particle filter algorithm based WLAN indoor tracking technology; further improve indoor positioning accuracy by optimizing particle filter algorithm.Location-based positioning technology usually works in two phases: off-line phase and on-line phase. During the off-line phase, the system tabulates the signal strength received from the access points, stores fingerprint information and establish the model; during the on-line phase, users make use of signal collecting terminal to obtain real-time signal strength, and by using a specific tracking algorithm to search in the location map, complete location estimation. Two optimal algorithms for particle filter algorithm are proposed in the paper: particle position and the interpolation-based tracking algorithms.Firstly, the paper analyzes the model of WLAN indoor tracking system. Through the analysis of traditional tracking algorithms, based on location technology and Kalman filter tracking algorithm, and the study of particle filter algorithm, we understand the advantages and disadvantages of the traditional tracking algorithms in indoor environment, and further proposed the necessity of applying particle filter algorithm in WLAN indoor environment. Simultaneously the basic principles of particle filter algorithm are also proposed.Secondly, through the study of the math model of particle filter based WLAN indoor location algorithm, this paper proposes the optimized methods, particle position interpolation-based tracking algorithm, which are able to give the solution to the problem of how to obtain signal strength of particle points. The former distributes signal strength to similar particle points, and the later obtains signal strength in particle points by interpolation of signal strength of reference points through spatial correlation. These two optimized algorithms can collect complete signal strength in particle points. According to histogram and method of kernel function, probability density distribution of particle points is gotten and location estimation of mobile terminals is implemented.Finally, we make use of indoor tracking algorithms and optimized particle filter algorithms to complete tracking through testing and simulation. By comparison and analysis of above algorithms, the effects of optimized algorithms are verified. And also different methods of tracking algorithms are provided for the program.