Study On Indoor Localization With WiFi

The indoor wireless location technology based on WiFi under different signal model is studied by Matlab simulation and experimental verification. The main content includes the following aspects:(1) The basic theory of wireless location technology is studied, including wireless distance measuring, common indoor location and factors that affect indoor location. The existing common path loss model, the applicable environment of different models and the relationship between path loss and distance have been discussed in this paper.(2) The location algorithms based on received signal strength(RSSI) are simulated to study their advantages and disadvantages. We simulate triangle centroid method, least squares method by Matlab, and use average location error and cumulative distribution function(CDF) to evaluate the location performance. The simulation results show that:the location performance of triangle centroid method is better than that of least squares method. Considering other factors that affect the location accuracy, we design experiments which adjust the position of the beacon nodes and transmit power, the results show that:the transmit power of beacon nodes has no influence on the location results basically; the location accuracy reduces as the range of the beacon nodes increases. So we should determine the number and location of beacon nodes according to the location area size.(3) The location algorithm based on the received signal strength is verified by a set of experments. Arranging APs in a laboratory environment, the received signal strength is scanned by the wireless network card integrated in the PC, and then we estimate the model parameters of logarithmic path loss model and multi-wall mode by minimum mean square error method(MMSE). Next we verify the location algorithm based on the fitting path loss model.In conclusion, based on the theoretical analysis, we simulate the location algorithms by Matlab, and then verify the location algorithm through measuring the received signal strength in a laboratory environment. The simulation results are in an acceptable error range, the error of measurement results are slightly larger than the simulation results, because the signal multipath effect and NLOS propagation have more impact on the location error in the actual environment than the ideal one.