Efficient Respiration Rate Estimation Based on MIMO mmWave Radar

Respiration rate provides important reference for disease diagnosis and rehabilitation treatment. MmWave-based respiration rate estimation has increasingly attracted attention in recent years due to its non-intrusiveness and cost-effectiveness. However, when the subject locates far away from the mmWave radar and also deviated from it, the low accuracy of respiration rate estimation becomes a major concern. This paper presents MMRes, a mmWave-based respiration monitoring system that can achieve high accuracy of respiration rate estimation even in the scenario of long range and large deviation angle. In order to extract desired information from weak signals, MMRes finds the most sensitive channel from all channels and extract respiration rate by adaptively decomposing the signals in the channel. Specifically, MMRes first selects diverse channels by a coarse screening method, then finds the best one with the maximum total energy by an optimal channel selection method, and finally decompose the signals of the best channel by a WOA-VMD based algorithm with optimized parameters. The computing time of MMRes is reduced by data selection in the coarse screening and parallel computation in the optimal channel selection. Our extensive experiments show that MMRes can improve the accuracy of respiration rate estimation by up to 60.5%, when the subject is 4 m away from the radar and the deviation angle is 60 \(^\circ \) C. The fast coarse screening can reduce the computing time by up to 76.2%, while the optimal channel selection with 8 cores can reduce the time by up to 33.7% than that with 4 cores.