Multi‐channel fingerprint localisation algorithm for wireless sensor network in multipath environment. Issue 15 (6th October 2017)

• Record Type: Journal Article
• Authors: Fang, Xuming; Nan, Lei; Jiang, Zonghua; Chen, Lijun
• Is Part Of: IET communications. Volume 11:Issue 15(2017)
• Page Start: 2253
• Page End: 2260
• Publication Date: 2017-10-06
• Subjects: wireless sensor networks -- wireless channels -- sensor placement -- RSSI -- adaptive Kalman filters; multichannel fingerprint localisation algorithm -- wireless sensor network -- WSN -- fingerprint localisation technology -- received signal strength indication -- RSSI -- measurement noise -- fingerprint positioning algorithms -- target–beacon distance -- target–reference distance -- adaptive Kalman filter -- memetic algorithm; Telecommunication systems -- Periodicals; Speech processing systems -- Periodicals; 621.38205
• DOI: 10.1049/iet-com.2017.0429
• Languages: English
• ISSNs: 1751-8628

Abstract

Abstract : Fingerprint localisation technology based on received signal strength indication (RSSI) is an active area of study in wireless sensor network (WSN) research, mainly because it is cheap and easy to implement. However, due to measurement noise and the multipath effect in RSSI, the accuracy of state‐of‐the‐art fingerprint positioning algorithms is diminished. To solve this problem, the authors propose a multi‐channel fingerprint localisation algorithm for WSNs in multipath environments. This algorithm also addresses the issue of existing techniques based on multi‐channel signal strength requiring an accurate initial estimate of target–beacon distance or prior knowledge of the target–reference distance. Their proposed algorithm first uses an adaptive Kalman filter to reduce the noise in RSSI measured in different channels, and then calculates the matched fingerprint according to the weight of different channels. Finally, a memetic algorithm is utilised to generate the optimised estimate of fingerprint and location. Extensive experimental results on an actual WSN testbed show that the proposed algorithm improves the accuracy of the existing fingerprint localisation algorithm by at least 50%, regardless of the placement of target node, the number of beacon nodes, and the number of calibration points.