Towards Radar-Agnostic Gait Analysis Across UWB and FMCW Systems

TL;DR

This work tackles cross-modality generalization in radar-based gait analysis by applying a single processing framework to collocated IR-UWB and FMCW radars. It introduces a modality-agnostic walking-segment detector and demonstrates that both radar types yield highly concordant spatiotemporal gait estimates with gold-standard motion capture, achieving mean accuracies in the 85–98% range and inter-modality differences below 4.1%. The results show near-identical trajectory fidelity, high gait-event detection accuracy, and strong cross-modality correlations (e.g., walking speed r≈0.98, ICC≈0.97), supporting the feasibility of radar-agnostic systems for scalable in-home gait monitoring. The findings pave the way for multi-device data fusion and portable models that generalize across radar modalities, with future work extending validation to home environments and clinical populations.

Abstract

Radar sensing has emerged in recent years as a promising solution for unobtrusive and continuous in-home gait monitoring. This study evaluates whether a unified processing framework can be applied to radar-based spatiotemporal gait analysis independent of radar modality. The framework is validated using collocated impulse-radio ultra-wideband (IR-UWB) and frequency-modulated continuous-wave (FMCW) radars under identical processing settings, without modality-specific tuning, during repeated overground walking trials with 10 healthy participants. A modality-independent approach for automatic walking-segment identification is also introduced to ensure fair and reproducible modality performance assessment. Clinically relevant spatiotemporal gait parameters, including stride time, stride length, walking speed, swing time, and stance time, extracted from each modality were compared against gold-standard motion capture reference estimates. Across all parameters, both radar modalities achieved comparably high mean estimation accuracy in the range of 85-98%, with inter-modality differences remaining below 4.1%, resulting in highly overlapping accuracy distributions. Correlation and Bland-Altman analyses revealed minimal bias, comparable limits of agreement, and strong agreement with reference estimates, while intraclass correlation analysis demonstrated high consistency between radar modalities. These findings indicate that no practically meaningful performance differences arise from radar modality when using a shared processing framework, supporting the feasibility of radar-agnostic gait analysis systems.

Figures (7)

• Figure 1: Experimental setup: (A) Participant walking along the designated 8-meter pathway, and (B) camera and radar sensor positions.
• Figure 2: Data processing pipeline for the UWB and FMCW radar systems.
• Figure 3: Examples of range profiles before and after preprocessing, and contrast-enhanced RDT frames, for the FMCW (a, b) and UWB (c, d) radars.
• Figure 4: Autocorrelation-based detection of walking segments. (a) Radar-derived feet-speed trajectory. (b) Autocorrelation-derived confidence level (left axis) and binary walking mask (right axis), where the black dashed line represents the 75% confidence threshold. (c) Feet-speed and (d) torso-range trajectory with detected walking intervals highlighted.
• Figure 5: (a) Mean gait parameter values across participants for FMCW and UWB radar systems, with shaded regions indicating ±1 standard deviation. (b) Distribution of gait parameter estimation accuracy for each radar system.