Research On RFID Indoor Location Algorithms Based On Scenario Analysis Method

In recent years, with the development of economy and the improvement ofpeople’s living standards, a lot of large-scale indoor spaces, such as superstore,underground parking, convention center, spring up, in which people have anincreasing demand for the location information. However, because of the complicatedinternal environment of these buildings, the multipath effect of the signal is apt to takeplace, which leads to mobile wireless positioning technology such as GPS as we haveknown can not meet people’s demand for the location accuracy; moreover, the closureof the building can seriously cripple the signal strength and even block the signal,which will result in the failure of location. So these technologies are not applicable inindoor environment.With the development of information technology and wireless communicationtechnology, researchers begin to use infrared, ultrasonic, bluetooth and zigbee in thefield of indoor localization. Of all the new technologies, RFID, because of its nlos,low cost, high precision and environment adaptiveness, has become the priorityselection. The main contents of this paper are as following：1）The paper introduces the classification, composition, operating principle andthe domestic and aboard development of RFID, and introduces three common indoorlocalization methods——Triangulation, scenario analysis method and K nearestneighbors. On that basis the author introduces TOA、TDOA、RSS、AOA andrespectively discusses the advantages and deficiencies of these algorithms. At last theauthor decides to use scenario analysis method as the main method for indoorlocalization.2）The paper introduces two kinds of classic RFID location algorithms based on scenario analysis method——LANDMARC and VIRE, the latter improved andadded VIVT on the basis of the former, and finds out the main error reason of VIREalgorithms by simulation experiments，comparison of results and error analysis. Errorreasons include signal fluctuations resulting from indoor mobility, unable toaccurately locate the point to be measured out of the reference tag, and the influenceof small possibility event.3）In view of error reasons, the paper comes up with three optimization methodsof algorithms, that is, to choose the mean value on the basis of multimetering, to addthe boundary virtual reference tag and to bring in the neighboring map, on the basis ofwhich the paper also puts forward a new algorithm——NEW-VIRE, in theexpectation of further improving location accuracy and decreasing errors withoutadding the system cost.4）The paper experiences with NEW-VIRE algorithms on MATLAB, verifies thelocation result, compares with the existed classic algorithms and discusses the mainparameter in the algorithm, such as the influence on location accuracy of the gridnumber of virtual reference tag and threshold value.5）The paper concludes the whole paper, points out deficiencies and has aprospect for the following research.