基于空间域压缩感知的车载雷达目标定位算法

Target positioning algorithm for automotive radar based on compressed sensing in spatial domain

车载毫米波雷达(MWAR)是高级驾驶辅助系统(ADAS)和自动驾驶系统(ADS)中一种重要的传感器。为了提高车载雷达的距离分辨率,降低对目标车的角度估算时对快拍数,该文构建了基于压缩感知的车载雷达目标定位算法。该方法从空间域角度进行分析,考虑车载雷达探测范围内目标在整个空间的稀疏分布,对目标的距离、速度和角度进行估算。相比传统车载雷达使用匹配滤波器无法区分相近目标,该压缩感知雷达可以区分邻近的目标。结果表明:在快拍数为1的情况下,本算法可以实现目标的角度估计,角度分辨率小于4°,该分辨率优于Music(多重信号分类)算法。

Millimeter wave automotive radar(MWAR) is an important sensors in the advanced driver assistance system(ADAS) and the automatic driving system(ADS). To promote the distance resolution by the MWAR,and to reduce the snapshot number when estimating the angle of the target vehicle, this paper proposed a positioning algorithm of automotive radar based on compressive sensing. The method analyzes from the perspective of spatial domain, considering the sparse distribution among the targets in the spatial domain of the whole space in the automotive radar detection range to estimate the distance, the velocity and the angle of the target vehicles. Compared to conventional vehicle-mounted radars, matched filters cannot distinguish similar targets. The compressed-sensing radar can distinguish adjacent targets. The results show that when a snapshot number is 1, the algorithm estimates the target angle with a resolution of less than 4°. The resolution is better than that by the Music(multiple signal classification) algorithm.