RSSI Based Positioning Fusion Algorithm in Wireless Sensor Network Using Factor Graph

Various positioning techniques have been widely developed based on received signal strength indicator (RSSI) in Wireless Sensor Network (WSN) positioning systems. Multilateration-based positioning technique is simple and easy to realize, but it can not provide very high positioning accuracy caused by fluctuation of range measurement. Fingerprinting technique is a promising method benefitting from its high precision. However, the process of building radio map cost too much time and labor. In this paper, a fusion algorithm based on both multilateration and fingerprinting is proposed to reduce cost and maintain high accuracy at the same time. An adaptive radio propagation mode is presented in this algorithm as well as a multilateration approaches based on sparse fingerprint. Factor graph is adopted to fuse the results of these two positioning techniques. Simulation experiments demonstrate that the proposed positioning fusion algorithm performs much better than any of the original algorithms participated in the fusion process.