FPNet: Joint Wi-Fi Beamforming Matrix Feedback and Anomaly-Aware Indoor Positioning

TL;DR

FPNet is presented, a unified deep learning framework that jointly addresses channel feedback compression, accurate indoor positioning, and robust anomaly detection (AD) and leverages the beamforming feedback matrix (BFM), a compressed CSI representation natively supported by IEEE 802.11ac/ax/be protocols.

Abstract

Channel State Information (CSI) provides a detailed description of the wireless channel and has been widely adopted for Wi-Fi sensing, particularly for high-precision indoor positioning. However, complete CSI is rarely available in real-world deployments due to hardware constraints and the high communication overhead required for feedback. Moreover, existing positioning models lack mechanisms to detect when users move outside their trained regions, leading to unreliable estimates in dynamic environments. In this paper, we present FPNet, a unified deep learning framework that jointly addresses channel feedback compression, accurate indoor positioning, and robust anomaly detection (AD). FPNet leverages the beamforming feedback matrix (BFM), a compressed CSI representation natively supported by IEEE 802.11ac/ax/be protocols, to minimize feedback overhead while preserving critical positioning features. To enhance reliability, we integrate ADBlock, a lightweight AD module trained on normal BFM samples, which identifies out-of-distribution scenarios when users exit predefined spatial regions. Experimental results using standard 2.4 GHz Wi-Fi hardware show that FPNet achieves positioning accuracy above 97% with only 100 feedback bits, boosts net throughput by up to 22.92%, and attains AD accuracy over 99% with a false alarm rate below 1.5%. These results demonstrate FPNet's ability to deliver efficient, accurate, and reliable indoor positioning on commodity Wi-Fi devices.

Figures (15)

• Figure 1: Channel sounding procedure in IEEE 802.11 Standard.
• Figure 2: An illustration of reconstruction-based AD using a trained autoencoder. Familiar CSI samples (top) result in low reconstruction error, while unfamiliar samples (bottom) yield high reconstruction error and are thus flagged as anomalies.
• Figure 3: Current CSI-based positioning framework where the reconstructed CSI is used for positioning.
• Figure 4: Illustration of the FPNet framework. The compressed BFM is used for both reconstruction and positioning.
• Figure 5: Overview of the FPNet framework for joint BFM feedback, positioning, and AD. At the STA, CSI $\mathbf{H}$ is decomposed into the BFM $\mathbf{V}$ via SVD, compressed into low-dimensional codewords, and then quantized into feedback bits. At the AP, the received bits are dequantized and processed by a decoder and a positioning network. The reconstructed BFM $\mathbf{V}'$ is also fed into ADBlock, which produces $\mathbf{V}'_{\mathrm{ad}}$. The MSE between $\mathbf{V}'$ and $\mathbf{V}'_{\mathrm{ad}}$ is compared to a threshold $\lambda$ to detect anomalies.