Realtime-Superlets Algorithm for Realtime Super-Resolution Heart Rate Estimation Using FMCW Radar

This paper presents a realtime-superlets (RT-SL) algorithm for monitoring heart rate (HR) in FMCW radar systems. The algorithm achieves realtime monitoring and super-resolution for accurate HR estimation. In the context of HR monitoring using FMCW radar, time-frequency analysis methods are essential for analysing time-varying HR signals. Conventional time-frequency analysis methods often yield ambiguous time-frequency representations. In comparison, the SL algorithm can generate super-resolution time-frequency maps compared to conventional methods. One limitation of the SL algorithm lies in the inability to process data in realtime. To address this limitation, we propose the RT-SL algorithm in this paper. The RT-SL algorithm can process data in realtime by right-aligning the data. We verify the effectiveness of the RT-SL algorithm using a 24 GHz FMCW radar. The experimental results show that the proposed RT-SL algorithm provides superior super-resolution in the HR time-frequency map. It outperforms traditional time-frequency analysis methods in HR estimation.