LFMCW 车载防撞雷达信号处理模块研究

摘要

For vehicle collision avoidance radar system,a low-complexity signal processing module is de-signed.It can make target detection and parameter estimation,and consists of several sub-modules comple-ting pre-processing,constant false alarm rate (CFAR)detection and multi-target pairing.Due to the hard-ware restriction,multiple transmitters is used to transmit signal in time diversity.In the pre-processing stage,phase calibration of signal echoes of different transmitters is made to form anequivalent virtual receiv-ing array with larger aperture,which results in more beamforming gain and higher angle estimation accuracy. In addition,digital beamforming is adopted to form narrow beams to suppress clutters.To decrease the effect of target shading and improve the detection rate,we propose an improved CFAR algorithm refreshing the samples used to estimate noise power adaptively.For multi-target pairing,we design a novel multistep pai-ring algorithm to remove false target by using priori knowledge to reduce pairing range.Simulation results demonstrate the effectiveness of our designed module.%针对 LFMCW 车载防撞雷达系统，设计了一种低复杂度的信号处理模块，该模块由预处理、恒虚警(CFAR)检测和多目标配对等子模块组成，能够实现目标检测与参数估计功能。由于系统硬件限制，多个天线不能同时发射信号，因此该系统通过时间分集发射信号。在预处理阶段，通过相位校准将接收的不同发射天线的目标回波信号合成一个具有更大孔径的等效虚拟接收阵列信号，获得更高的波束形成增益和测角精度，并对虚拟接收阵列信号采用数字波束形成技术形成窄波束接收以抑制杂波干扰信号。在 CFAR检测中，提出了一种能够自适应改变噪声电平估计样本的改进 CFAR 算法以有效减小目标遮蔽效应并提高检测概率。针对多目标配对问题，提出了一种利用先验信息压缩匹配空间的多步配对算法，能够有效去除虚假目标，并降低配对复杂度。仿真结果表明，该模块目标检测成功率较高，且能够对速度、距离以及角度进行有效估计。