应用于超宽带呼吸检测的压缩感知算法研究

超宽带雷达是最近生命探测采用的主要形式,绝对带宽不小于500 MHz的带宽对前端AD采样速率要求较高。针对超宽带系统硬件成本很高有时甚至不能实现的系统搭建问题,论文提出自适应的压缩感知(CS)呼吸检测算法。根据压缩感知欠采样的特点,首次将CS原理应用在微多普勒呼吸频率检测中,对OMP和GPSR两种恢复算法进行了测试,最终采用了OMP检测算法,并设计了自适应观测字典。应用本方法采样率可以达到0.66,大大降低了前端采样压力,节约硬件实现成本,并提高了检测效率。

Research on compressed sensing applied to UWB respiratory detection

Ultra-wideband (UWB) radar is the main form of the recently adopted life detection, but AD of fore-end absolute bandwidth of not less than 500 MHz requires a higher sampling rate. Self-adaptive compressed sensing (CS) breath detection algorithm is proposed to address the high UWB hardware cost and failure of implement sometimes. According to the characteristics of under-sampling of compressed sensing, CS principle is firstly applied in the micro-Doppler respiratory frequency detection. OMP and GPSR recovery algorithms are tested and the OMP is finally chosen with the design of self-adaptive observation dictionary. The sampling rate is reduced to 0.66 with the method, which can greatly reduce the fore-end sampling pressure, save hardware cost of implement, and improve detection efficiency.