Publications

CONFERENCE (INTERNATIONAL) MIMO-Free AoA Localization for Uncooperative Wi-Fi Probe Requests Using Commodity Single-Input Receivers

Shiyuan Zhuang (Institute of Science Tokyo), Kota Tsubouchi, Nobuhiko Nishio (Ritsumeikan University), Masamichi Shimosaka (Institute of Science Tokyo)

The 16th International Conference on Indoor Positioning and Indoor Navigation (IPIN 2026)

November 18, 2026

Passive localization of Wi-Fi probe request (PR) frames is important for crowdsensing and pedestrian flow analysis in uncooperative indoor environments. Existing passive approaches mainly rely on RSSI- or CSI-based reverse fingerprinting, which requires environment-specific calibration and maintenance. Such dependence limits scalability and robustness to environmental changes. In this paper, we propose a MIMO-free, fingerprint-free localization method for PR frames using angle-of-arrival (AoA) estimation from CSI. Our method aligns PR frames across commodity single-input receivers, processes CSI phase and amplitude features, filters corrupted frames, and estimates AoA with a lightweight 1D convolutional network. The estimated AoAs are then combined geometrically to localize the signal source. Experiments in a real indoor environment show that the proposed method outperforms an RSSI-based passive baseline with only a small number of receivers. These results demonstrate that accurate uncooperative localization is feasible without dense fingerprint databases or specialized MIMO hardware.

Paper : MIMO-Free AoA Localization for Uncooperative Wi-Fi Probe Requests Using Commodity Single-Input Receiversopen into new tab or window (external link)