A Study Of Indoor Postioning Algorithm Base On WLAN And System Realization

In recent years, with the rapid development of mobile Internet technology, Location Based Services(LBS) have been widely applied in our life, such as the traffic navigation, store navigation, logistics management and etc. In addition, LBS gradually expand from outdoor to indoor. To a get good indoor experience for LBS, there must be a qualified and reliable indoor positioning system. While,restricted by the cost, the existing indoor positioning system could not be widely applied. However, cheap indoor positioning now can be available with the popularity of wireless networks and based on the application of existing wireless networks, the indoor positioning cost is reduced greatly without the loss of the performance. So, Wireless Local Area Network(WLAN) based positioning has been a hotpot in the research of indoor positioning technology for the advantages of its flexibility and low construction cost.There are two kinds of popular indoor wireless positioning methods, i.e., wireless-channel-model-based positioning and RF Fingerprinting method. In the fingerprinting method, a lot of offline gathering needs to be done and the RSSI should be gathered again once the environment is changed. This restricts its wide applications. Meanwhile, the ranging-based positioning method can get similar accuracy and can be realized by the existing wireless devices without any extra devices. Therefore, the cost is greatly saved which satisfies the LBS. Thus, this thesis mainly researches on the ranging-based positioning algorithm.Firstly, this thesis gives an introduction of the existing indoor positioning system. And then, researches on the wireless channel model indoors are done comparing with the measured the data so that the parameters of the model can be certained. After that, this thesis goes on giving a introduction of the existing ranging-based indoor positioning algorithms, such as the nonlinear least squares method(NLS), linear least squares method(LLS), maximum likelihood algorithm(MLE). With the analysis of the problem of the existing algorithms, the thesis shows an improved method, weighted least squares method(WLS). From the perspective of a set of evaluation metrics including accuracy, precision, robustness, complexity, and stability, this thesis compares the four methods and finds that the WLS gives better performance. And then, a lot of work is being done to gets a better weight coefficient. Furthermore, the positioning system is implemented which is based on the WLS and at the same time, a lot of improvement is done. The running results of the positioning system show that WLS has high positioning accuracy, low complexity and its cost is low. So, WLS is a suitable method for indoor positioning.