| 摘要: |
| 多输入多输出调频连续波（MIMO-FMCW）雷达具备距离与角度分辨能力，能够有效实现多目标生命体征监测。针对多目标生命体征监测中易受目标混叠与相互干扰的问题，本文提出了一种结合二维多重信号分类（MUSIC）算法定位与角域约束多零陷波束形成的多目标生命体征监测方法。首先，基于多通道回波数据构造样本协方差矩阵，并通过空间平滑预处理进行解相干。利用最小描述长度（MDL）准则估计目标数目，进而采用二维MUSIC算法实现距离和方位角的联合估计。然后，将目标定位结果作为角域约束多零陷波束形成的输入，在保持目标输出无失真响应的前提下，通过在其余目标邻近区域内设置多个零陷以抑制多目标间的互扰，从而提取各目标的慢时间维信号。最后，对波束形成输出的慢时间维信号进行相位解缠绕与平滑处理，并采用变分模态分解（VMD）重构呼吸与心跳信号。实验结果表明，所提方法在多目标场景下可实现较低的呼吸率与心率平均相对误差，显著提升了生命体征参数的估计精度。 |
| 关键词: 生命体征监测 调频连续波雷达 多重信号分类 波束形成 |
| DOI： |
| 分类号:TN957.52 |
| 基金项目:辽宁省科技计划联合计划基金项目（2025110387-JH2/1018） 辽宁省高校基本科研业务费项目（LJ222410143071） |
|
| Multi-Target Vital Sign Monitoring Based on Beamforming for MIMO-FMCW Radar |
|
|
|
|
| Abstract: |
| Multiple-input multiple-output frequency-modulated continuous-wave (MIMO-FMCW) radar provides range and angular resolution, enabling effective multi-target vital sign monitoring. To address target aliasing and mutual interference in multi-target vital sign monitoring, this paper proposes a multi-target vital sign monitoring method that combines two-dimensional multiple signal classification (2D-MUSIC) localization with angularly constrained multi-null beamforming. First, a sample covariance matrix is constructed from multi-channel echo data, and spatial smoothing is applied for decorrelation. The minimum description length (MDL) criterion is then used to estimate the number of targets, and the 2D-MUSIC algorithm is employed to jointly estimate the range and azimuth of each target. Next, the localization results are used as the input to the angularly constrained multi-null beamforming module. While maintaining a distortionless response toward the desired target, multiple nulls are imposed within the angular neighborhoods of the other targets to suppress mutual interference, thereby extracting the slow-time signal of each target. Finally, phase unwrapping and smoothing are performed on the beamformed slow-time signals, and variational mode decomposition (VMD) is used to reconstruct the respiration and heartbeat signals. Experimental results show that the proposed method achieves low average relative errors in respiration rate and heart rate estimation in multi-target scenarios, significantly improving the accuracy of vital sign parameter estimation. |
| Key words: vital sign monitoring frequency modulated continuous wave radar multiple signal classification beamforming |