Super-Range-Resolution Respiratory Signals Detection From Multiple Targets With Single-Channel FMCW Radar

In recent years, multiperson respiration detection has gained increasing attention, especially in applications where multiple individuals need to be monitored simultaneously, such as in healthcare or smart home environments. However, accurately detecting the vital signs of multiple people using single-channel radar remains a significant challenge due to the complexity of signal separation in overlapping or closely spaced targets. This article presents a novel complex mix source separation (MSS) model for multiperson respiration detection using single-input-single-output (SISO) radar. The proposed model enables the separation of respiratory signals from multiple individuals without the need for additional antennas. A detailed analysis of the model’s conditions is provided, demonstrating its ability to separate multiperson respiratory signals even when the separation distance between individuals is far below the radar’s range resolution. An improved Nesterov complex entropy bound minimization (Nesterov-CEBM) algorithm is proposed, which enables multiperson vital sign signal separation in the complex domain, avoiding the nonlinearities of traditional methods and significantly improving the computational efficiency. Comprehensive experiments have been conducted to demonstrate that the proposed approach successfully separates the respiratory signals of two closely spaced individuals, with an average respiratory rate error of no more than 0.5 bpm. The proposed approach does not require sophisticated hardware systems, making it particularly suitable for practical applications such as sleep monitoring and patient care.