Experimental analysis of RSSI-based localization algorithms with NLOS pre-mitigation for IoT applications
Fabrizio Carpi,Marco Martalò,Luca Davoli,Antonio Cilfone,Yingjie Yu,Yi Wang,Gianluigi Ferrari
DOI: https://doi.org/10.1016/j.comnet.2023.109663
IF: 5.493
2023-03-04
Computer Networks
Abstract:In this paper, we propose an effective target localization strategy for Internet of Things (IoT) scenarios, where positioning is performed by resource-constrained devices. Target-anchor links may be impaired by Non-Line-Of-Sight (NLOS) communication conditions. In order to derive a feasible IoT-oriented positioning strategy, we rely on the acquisition, at the target, of a sequence of consecutive measurements of the Received Signal Strength Indicator (RSSI) of the wireless signals transmitted by the anchors. We then consider a pragmatic approach according to which the NLOS channels are pre-mitigated and "transformed" into equivalent Line-Of-Sight (LOS) channels to estimate more accurately each target-anchor distance. The estimated distances feed "agnostic" localization algorithms, operating as if all links were LOS. We experimentally assess the performance of our approach in indoor (IEEE 802.11-based) and outdoor (Long Term Evolution, LTE-based) scenarios, considering both geometric and Particle Swarm Optimization (PSO)-based localization algorithms. Even if NLOS mitigation per single communication link is very effective, our results show that, in a given environment, it is possible to derive an "average" NLOS mitigation strategy regardless of the specific position of the target in the given environment. This is crucial to limit the computational complexity at IoT nodes performing localization, yet guaranteeing a relatively high (for IoT scenarios) localization accuracy, especially in an IEEE 802.11-based indoor case (with six anchors). The obtained performance compares favorably (in relative terms) with that obtained with more sophisticated wireless technologies (e.g., Ultra-WideBand, UWB).
computer science, information systems,telecommunications,engineering, electrical & electronic, hardware & architecture