An adaptive localization algorithm based on RSSI in wireless sensor networks

Abstract

Wireless sensor networks (WSN) have been used in various application areas and will be more prosperous in the future. The technology of localization, by which the position of blind node can be acquired, is a significant factor and one of the supporting technologies in WSN. Most of the existing localization algorithms in sensor networks mainly depend on one simple model and can only be used in a given condition, which cannot be adapted to the complex environment and thus restrict their application. In this paper, a novel adaptive scheme, based on the relationship between the RSSI (Received Signal Strength Indication) values and the spread distance of RF signal, is presented by fusing the experience model and the theoretical model in traditional RSSI-Based localization methods for improving the adaptability to different environments and enhances the pointing accuracy in WSN. The key method of this scheme is that through dividing the entire testing set into several pieces and placing characteristic node in every zone, one of the two models will be dynamically selected after comparing the error between the real distance the RF signal traveled and the distance computed from each of the two models by the characteristic node to minimize the positioning error.