Human movement effects on the performance of the RSSI-based trilateration method: adaptive filters for distance compensation

• Published in: Journal of reliable intelligent environments Vol. 6; no. 2; pp. 67 - 78
• Main Authors: Sasiwat, Yoschanin, Buranapanichkit, Dujdow, Chetpattananondh, Kanadit, Sengchuai, Kiattisak, Jindapetch, Nattha, Booranawong, Apidet
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.06.2020; Springer Nature B.V
• Subjects: Adaptive filters; Artificial Intelligence; Computer Science; Errors; Health Informatics; Human motion; Original Article; Performance and Reliability; Signal strength; Simulation and Modeling; Software Engineering/Programming and Operating Systems; Trilateration; User Interfaces and Human Computer Interaction; Wireless networks; Trilateration; Human movements; Indoor; RSSI-based localization; Experiment
• ISSN: 2199-4668; 2199-4676
• DOI: 10.1007/s40860-019-00094-x

In this paper, an experimental study of human movement effects in indoor wireless networks on the performance of the received signal strength indicator (RSSI)-based trilateration method is presented. The contribution of this work is that how the RSSI fluctuation caused by human movements nearby a target node influences the estimation accuracy of the trilateration method is studied. Additionally, efficient adaptive filters which can observe RSSI variation levels and select appropriate inputs (i.e., distance values converted from RSSI values) are proposed for handling the RSSI variation problem and compensating the position estimation error. Experiments using low-cost 2.4-GHz wireless nodes have been set and tested in a laboratory room. Results demonstrate that, during the human blocking the communication link between the transmitter and the receiver, RSSI data are fluctuated and the trilateration method shows a large estimation error. Here, the localization accuracy significantly depends on the RSSI variation level. The experimental results also indicate that, by applying our proposed solutions, they can automatically provide appropriate inputs for the trilateration method. The estimation error decreases by 43.76% and 46.28% for our test scenarios.