Vital Sign Tracking Using Modified Constant Modulus Algorithm Adaptive Array

Abstract

This paper proposes a modified constant modulus algorithm (CMA)-based beamforming method for a multiple-input multiple-output radar to track and detect vital signs via blind processing. Our method consists of two key functions: (1) the beamforming weight and DC offset are jointly updated to remove the semi-constant clutter and maintain the constant amplitude, which is the original function of CMA, and (2) the weight is updated by considering the phase variance owing to the typical respiratory response to make the weight converge to the vital signs. The proposed method was experimentally evaluated by comparing its heart rate estimation accuracy with that of the conventional CMA-based and maximal-ratio combining (MRC)-based beamforming methods. The experimental results showed that the proposed beamforming method with and without the phase variance constraint achieved the median errors of 3.45 and 4.06 bpm from the reference electrocardiogram, respectively, whereas the conventional CMA-based and MRC-based beamforming methods achieved the median errors of 21.1 and 9.79 bpm, thereby demonstrating the effectiveness of the proposed method.