Study On Indoor Wifi Localization

With the development of economy, the demands of position information becomes bigger and bigger, and the LBS (location based service) has attract more and more attention in the industry and academia. "Location services is also called the location-based services. It is a value-added services provided by communication network and satellite positioning system (GPS) together,which send the information(such as latitude and longitude coordinates) of mobile through a set of positioning technology to the users or others communication system, and then realize various position related business" In essence, it is a kind of new service business, which is more general and related to the space position. Three goals (3w) of LBS had been put forward by American scholar Schilit. First where you are (spatial information). Second, the people who you are with (social information), Third, around(query information) which have constituted foundation of the geographical location of the serviceLBS is divided into indoor and outdoor location services location services, outdoor positioning technology originated in modern military technology, modern military dependence on the satellite has reached an unprecedented height, From reconnaissance, early warning, remote sensing, surveillance, command and communications, and the precision-guided are inseparable from the satellite technology to provide positioning technology, the United States first to set up the GPS project in 1958, and put into use in 1964.As so far, the outdoor location system has four mature system, respectively for the GPS (Global Position System) system, the Russian GLONASS (GLONASS) system, the EU’s Galileo system, China’s Beidou system.At present, WiFi indoor positioning technology is facing two major difficulties, the first is the data acquisition, the second is the robustness problem. This paper from a new perspective, the researcher adopts Naive Bayesian algorithm and support vector machine algorithm for indoor positioning, and achieved good results.We propose a new algorithm. First the area is divided into regions, and then divided into points. Using some strong signal to decide the region by Naive Bayes algorithm which relax the accuracy, then filter signal and use the support vector machine algorithm to decide the point.The advantages:1, Fast. In larger area, the speed of our algorithm will be improved significantly.2, High precision. Compared with the traditional support vector machine we can improve the accuracy of 4%.3, Robustness. With respect to the naive Bayes algorithm, we conducted a filtered signal, and using the support vector machine, increasing the robustness of the algorithm.