A Multi-Target Localization and Vital Sign Detection Method Using Ultra-Wide Band Radar

Area of Science:

• Engineering
• Biomedical Engineering
• Signal Processing

Background:

• Ultra-wideband (UWB) radar offers non-contact life detection for disaster rescue.
• Current multi-target detection methods struggle with accuracy and simultaneous vital sign extraction.
• There is a need for improved UWB radar techniques for locating survivors and monitoring their health.

Purpose of the Study:

• To propose a novel UWB radar method for multi-target localization and vital sign detection.
• To enhance accuracy in identifying and locating multiple individuals in complex environments.
• To simultaneously extract and analyze breathing and heartbeat information for individual targets.

Main Methods:

• Target recognition and localization using permutation entropy (PE) and K-means++ clustering.
• Adaptive denoising for vital sign extraction via ensemble empirical mode decomposition (EEMD) and wavelet analysis (WA).
• Heartbeat frequency extraction employing particle swarm optimization (PSO) and stochastic resonance (SR).

Main Results:

• The PE-K-means++ method achieved an average relative error of 1.83% for multi-target localization.
• The EEMD-WA method demonstrated an average relative error of 4.27% for reconstructed respiratory signal frequency.
• The PSO-SR method yielded an average relative error of 6.23% for heartbeat frequency detection.

Conclusions:

• The proposed UWB radar method effectively recognizes and locates multiple human targets.
• Accurate vital signs (breathing and heartbeat) are successfully extracted, providing crucial information for rescue.
• This technology offers a theoretical basis and technical support for emergency rescue systems.