A Framework Using Fingerprinting for Signal Overlapping-Based Method in WLAN

Abstract

In this paper, we proposed a novel method that contains the advantages of two kinds of RSSI-based localization methods, the RSSI Triangulation and the RSSI Fingerprinting methods, to improve some disadvantages of traditional RSSI-based localization methods. The advantage of the former is that we could determine the possible region of the wireless device without collecting a large amount of training data and making a great training effort. However, in most case Triangulation method cannot determine the position of the user precisely and usually leads to erroneous prediction by large distance. As for the latter, RSSI fingerprinting method wastes considerable amount of time and manpower to collect all possible RSSI in all possible regions. On the contrary, it always gets more precise result. This paper combines the advantage of these two schemes. The proposed scheme does not have to pre-collect large amount of RSSI records and makes use of the results of signal coverage overlapping method as the input of RSSI fingerprinting method in the online phase.