Abstract
The increasing demand for the Internet of Things (IoT) applications has driven the need for precise and scalable indoor positioning systems (IPSs). Among existing techniques, wireless fidelity (Wi-Fi) round-trip time (RTT), introduced by the IEEE 802.11mc standard, offers high accuracy but suffers from scalability limitations due to the frames specifically sent for localization purposes. While fingerprinting can improve system scalability compared to multilateration, it relies on computationally intensive machine learning algorithms, which can be challenging to deploy on resource-constrained IoT devices. This article proposes a novel approach to improve the scalability and efficiency of Wi-Fi RTT fingerprinting solutions. The proposed system aims to simplify clustering and feature selection tasks by sorting received-signal strength indicator (RSSI) measurements, thus eliminating the need for more complex algorithms. Experimental evaluations, in scenarios with varying densities of access points (APs), demonstrate significant performance gains compared to state-of-the-art (SoA) approaches. In high-density scenarios (e.g., one AP per 22 m2), the proposed system achieves a root-mean-squared error (RMSE) below 50 cm, a decrease of almost 98% in the memory used by the device, and supports location requests of up to 124 simultaneous Wi-Fi users, representing a 7.8% increase in scalability compared to previous proposals. These results highlight the potential of the proposed system to achieve a good compromise among accuracy, scalability, and complexity, making it a viable solution for IPS implementation in IoT environments.
Funding Information
  • Recovery, Transformation, and Resilience Plan, funded by European Union (Next Generation), under the Auspices of National Cybersecurity Institute (INCIBE) Cybersecurity Chair named CARISMATICA
  • Generalitat de Catalunya (2021-SGR-00594)