Study On The Range-based Localization Algorithm In WiFi Network

In recent years, with the spurt development of mobile terminals, especially thesmartphone as a symbol, as well as the need for the construction of an informatizationand intelligent city, wireless LAN technology also reaching its climax on the way ofhigh-speed development. WiFi hotspots can be found everywhere, whether it’s families,schools, office buildings, or shopping malls, airports, public traffic systems. At the sametime, location services become more and more popular, so that wireless locationtechnology ushered in a wave of its research and development boom. The thesis beginswith the introduction of wireless location technologies and systems, and then transitionto the positioning technologies in the WiFi network. It focus on the research oflocalization algorithms based on RSSI ranging, and the design and implement processof its application in the positioning system based on WiFi networks.Because of the conventional positioning systems based on RSSI technology, whichin its first stage of positioning, according to the function relationship between RSSI anddistance to establish an equation. Before the implementation of localization algorithms,it’s must to measure and processing a mount of data for calculate the parameters ofenvironment and hardwares. In the thesis, according to the mathematical principle thatthe trajectory of a point who’s constant distances ratio from two fixed points is a coniccurve. We design a "trajectory" method, which is a function between the coordinates ofthe measured point and the difference of two RSSI which come from two differentreference points. The relativization difference of two RSSI avoid to measure someparameters, it also avoid to calculation the distance of the equation itself. In the secondstage of positioning, according to the equations of the first phase to calculate thecoordinates of the measured point. The thesis describes a algorithm about the minimumsum of distances to the boundary, and draw a problem of nonlinear programming fromit, then calculate the coordinates of the measured point using iterative method.No matter how better and how complex about the algorithm design, we are mostconcerned about is the final positioning precision, it decided mainly by the accuracyRSSI measurements, except related with algorithms. Because the process of measuring the values of RSSI which directly effects the accurate result and precise position iseasily interfered by the ambient noise, thus causing the ranging error.So the introductionof several kinds of filter, to filter the random gauss noise in the processing of the RSSIdata acquisition, and smoothing the RSSI value. This filter processing can effectivelyreduce the interference of some noise, improve the positioning accuracy.Based on the above algorithm principles, we design a verification system based onWiFi network technology. The system based on the network-side which mainly aims atthe design and implementation of positioning service, which afford the core functionabout communication module, data processing module, and positioning engine, and alsofeedback the results to clients. Finally, the performance for the positioning system of variousalgorithms were analyzed and compared.