Enhancing Passive WiFi Device Localization Through Packet Timing Analysis

Abstract

This article presents an innovative timing-based localization method aimed at determining the positions of active WiFi devices through passive reception. The method involves capturing and analyzing the timing of over-the-air WiFi packets: request-to-send (RTS), clear-to-send (CTS), data (DATA), and acknowledgment (ACK) packets exchanged between WiFi routers and clients. The accuracy of round-trip time (RTT) estimation, crucial for distance calculation, can be affected by factors, such as clock variations between devices and, notably, the short interframe space (SIFS) time setting in the WiFi protocol. Despite SIFS time aiming to ensure a consistent interval between DATA and ACK frame transmissions, IEEE 802.11 standards permit up to a 10% variation in SIFS time. When combined with device-level disparities and environmental fluctuations, individual RTT measurements may not reliably estimate distances. In this study, we employ statistical clustering techniques, specifically k-means clustering, to enhance RTT estimation by refining coarse- and fine-timing estimates. Each captured packet pair, i.e., (DATA/ACK), is assigned to the cluster with the most similar coarse and fine RTT characteristics. Subsequently, the properties of the identified cluster (e.g., coarse RTT/fine RTT) are utilized as a more precise RTT estimate for localization computations. Simulations and experiments conducted under diverse multipath conditions demonstrate the algorithm’s accuracy in 2-D positioning, achieving an average accuracy of as low as 0.24 m in simulations and 1.18 m in experiments when the Wi-Fi router and device are separated by distances of up to 18 m. The proposed method offers a robust approach for accurate passive Wi-Fi positioning, highlighting its potential for real-world applications.