Doppler Radiance Field-Guided Antenna Selection for Improved Generalization in Multi-Antenna Wi-Fi-based Human Activity Recognition

2025-09-18

AI Summary

Wi-Fi channel state information (CSI) for human activity recognition (HAR) suffers from Doppler radiation field (DoRF) velocity projection distortion due to asynchronous access point (AP) clocks and environmental/hardware noise, severely degrading model generalization. Method: We propose a dynamic antenna selection and noise suppression framework grounded in DoRF fitting error. Specifically, we introduce DoRF fitting residuals as a novel criterion for antenna selection and jointly leverage CSI to construct a 3D motion representation—mitigating clock offset and additive interference directly at the signal level. Contribution/Results: The method requires no additional hardware or explicit synchronization mechanisms. It significantly enhances model robustness and cross-scenario generalization. Evaluated on a small-scale gesture dataset, it achieves substantial improvements in both classification accuracy and generalization performance, validating its effectiveness in realistic, complex environments.

Technology Category

• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Intelligent Robots: State Estimation
• Computer Vision: Motion & Tracking

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

With the IEEE 802.11bf Task Group introducing amendments to the WLAN standard for advanced sensing, interest in using Wi-Fi Channel State Information (CSI) for remote sensing has surged. Recent findings indicate that learning a unified three-dimensional motion representation through Doppler Radiance Fields (DoRFs) derived from CSI significantly improves the generalization capabilities of Wi-Fi-based human activity recognition (HAR). Despite this progress, CSI signals remain affected by asynchronous access point (AP) clocks and additive noise from environmental and hardware sources. Consequently, even with existing preprocessing techniques, both the CSI data and Doppler velocity projections used in DoRFs are still susceptible to noise and outliers, limiting HAR performance. To address this challenge, we propose a novel framework for multi-antenna APs to suppress noise and identify the most informative antennas based on DoRF fitting errors, which capture inconsistencies among Doppler velocity projections. Experimental results on a challenging small-scale hand gesture recognition dataset demonstrate that the proposed DoRF-guided Wi-Fi-based HAR approach significantly improves generalization capability, paving the way for robust real-world sensing deployments.

Problem

• Improving Wi-Fi HAR generalization via antenna selection
• Mitigating noise and outliers in CSI and Doppler data
• Enhancing robustness in multi-antenna sensing deployments

Innovation

• DoRF-guided antenna selection
• Suppress noise via fitting errors
• Improve Wi-Fi HAR generalization