Towards Heart Rate Estimation in Complex Multi-Target Scenarios: A High-Precision FMCW Radar Scheme Integrating HDBS...

Area of Science:

• Biomedical Engineering
• Radar Systems
• Signal Processing

Background:

• Non-contact heart rate estimation using frequency-modulated continuous wave (FMCW) radar is established for single targets.
• Multi-target scenarios present significant challenges in target discrimination and accurate heart rate estimation.
• Existing methods are underexplored in complex, multi-person environments.

Purpose of the Study:

• To develop a novel scheme for robust multi-target heart rate estimation using FMCW radar.
• To address the challenges of accurate target discrimination and vital sign extraction in crowded environments.
• To achieve high-precision, super-resolution heart rate estimation with reduced computational load.

Main Methods:

• High-precision distance-bin selection (HDBS) for range domain localization.
• Multiple-input multiple-output (MIMO) array processing with Root-multiple signal classification (Root-MUSIC) for angular domain discrimination.
• Variational mode decomposition (VMD), local mean decomposition (LMD), and wavelet thresholding (WT) cascade (VLW) for signal extraction.
• Improved fast iterative interpolated beamforming (FIIB) for efficient heart rate estimation.

Main Results:

• The proposed scheme effectively discriminates multiple targets in all tested scenarios, including dense arrangements and occlusion.
• Heart rate estimation achieved a mean absolute error (MAE) below 2.6 beats per minute (bpm).
• The improved FIIB algorithm demonstrated a ~60% reduction in execution time compared to the standard version.
• Sufficient signal-to-noise ratio (SNR) gain was achieved through low-complexity accumulation in distance and angle estimation.

Conclusions:

• The developed scheme offers a viable and robust solution for multi-target heart rate estimation.
• The integration of advanced signal processing techniques enables accurate vital sign extraction in complex environments.
• This technology has potential applications in various engineering fields requiring non-contact physiological monitoring.