Low-Complexity Offline and Online Strategies for Wi-Fi Fingerprinting Indoor Positioning Systems

Abstract Received signal strength (RSS) fingerprinting is a common solution for the implementation of Wi-Fi indoor positioning systems (IPSs). The trade-off between accuracy and complexity of this approach mainly depends on (1) a careful planning of the offline phase, during which RSS values from Wi-Fi access points are collected at predefined reference points, and (2) an optimized definition of the algorithm used in the online phase, during which position is estimated. This chapter analyzes and discusses the above trade-off, and identifies two low-complexity strategies for both phases. As regard the offline phase, RSS prediction, via multiwall multifloor indoor propagation model, is proposed, while an Affinity Propagation two-step algorithm is tested in the online phase. Experimental results show the feasibility of the proposed schemes, and how both strategies lead to a simpler system implementation, with respect to traditional schemes, while preserving positioning accuracy.