WiCAL: Accurate Wi-Fi-Based 3D Localization Enabled by Collaborative Antenna Arrays

TL;DR

WiCAL tackles the challenge of accurate 3D localization in Wi‑Fi by enabling multistatic localization through collaborative antenna arrays that leverage wavelength‑scale intra‑array coherence and inter‑array cooperation via switched URAs. The framework combines a three‑stage phase calibration, an I‑SSMUSIC AoA estimator with forward/backward spatial smoothing, a robust closest geometric point method for 3D positioning, and a direct position determination (DPD) approach that fuses data across distributed URAs. Hardware validation on commodity Wi‑Fi hardware demonstrates sub‑centimeter to decimeter accuracy in indoor environments, with median 3D localization errors as low as 15.6 cm using two URAs and precise AoA estimates (≈$1^ obreakspace^ gtr 1^{\circ}$). The results show WiCAL’s potential to enable scalable, accurate 3D positioning in next‑generation Wi‑Fi sensing and communication systems without requiring expensive multi‑transceiver modules.

Abstract

Accurate 3D localization is essential for realizing advanced sensing functionalities in next-generation Wi-Fi communication systems. This study investigates the potential of multistatic localization in Wi-Fi networks through the deployment of multiple cooperative antenna arrays. The collaborative gain offered by these arrays is twofold: (i) intra-array coherent gain at the wavelength scale among antenna elements, and (ii) inter-array cooperative gain across arrays. To evaluate the feasibility and performance of this approach, we develop WiCAL (Wi-Fi Collaborative Antenna Localization), a system built upon commercial Wi-Fi infrastructure equipped with uniform rectangular arrays. These arrays are driven by multiplexing embedded radio frequency chains available in standard access points or user devices, thereby eliminating the need for sophisticated, costly, and power-hungry multi-transceiver modules typically required in multiple-input and multiple-output systems. To address phase offsets introduced by RF chain multiplexing, we propose a three-stage, fine-grained phase alignment scheme to synchronize signals across antenna elements within each array. A bidirectional spatial smoothing MUSIC algorithm is employed to estimate angles of arrival (AoAs) and mitigate performance degradation caused by correlated interference. To further exploit inter-array cooperative gain, we elaborate on the synchronization mechanism among distributed URAs, which enables direct position determination by bypassing intermediate angle estimation. Once synchronized, the distributed URAs effectively form a virtual large-scale array, significantly enhancing spatial resolution and localization accuracy.

Figures (22)

• Figure 1: The system overview of WiCAL.
• Figure 2: Switched URA. Each port of the RF switches is connected to a unique antenna, with every trio of antennas grouped together to capture synchronized CSI data through Wi-Fi device. Each group is marked with the same color. For instance, by employing three SP4T RF switches, we can collect CSI data from a total of 12 antennas, organized into 4 distinct groups.
• Figure 3: (a) The signal source is placed in front of the antenna array, within the far-field, ensuring that $\tau^{p}$ remains consistent across all antennas. (b) The CPAs for inter-group calibration includes antennas {1, 6, 8, 10} and {13, 18, 20, 22}, enclosed in yellow boxes. Antennas {1, 6, 20}, marked in red, are used for phase alignment between arrays.
• Figure 4: Phase calibration and alignment. (a) The intra-group calibration eliminates the phase offsets introduced by radio chains. (b) To synthesize a virtually phase-synchronized CSI from individual CSI measurements, we use the inter-group synchronization to align the phase shifts caused by CFO.
• Figure 5: URA consisting of $M_x \times M_y$ antennas, where $M_x$ and $M_y$ denote the number of antennas along the x and y axes.