低比特量化阵列快速无网格化DOA估计方法  

A Fast Gridless DOA Estimation Algorithm for Low-bit Quantized Array

作　　者：夏钰 席峰 陈胜垚[1] XIA Yu;XI Feng;CHEN Shengyao(School of Electronic and Optical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南京理工大学电子工程与光电技术学院,南京210094

出　　处：《现代雷达》2022年第8期14-19,共6页Modern Radar

摘　　要：针对如何在低比特量化下实现高精度无网格化到达角(DOA)估计的问题,提出了一种基于原子范数最小化的快速DOA估计算法。该方法首先建立阵列接收信号低比特量化前后数据一致性的关系模型,然后采用原子范数最小化实现无网格化DOA估计,因此可将低比特量化下的无网格化DOA估计问题转化为求解量化数据与未量化数据的失配度和相应的原子范数联合最小化问题。为了高效求解该优化问题,该算法采用加速近端梯度技术,通过迭代地执行动量梯度下降和近端投影使得问题收敛。相比于传统基于特征子空间的算法,文中提出的算法在快拍数不足的情况下也同样适用。仿真结果表明:文中所提算法兼具估计的准确性和计算的高效性。Focusing on developing high accuracy gridless direction-of-arrival(DOA)estimation for low-bit quantized array,a fast gridless DOA estimation algorithm based on atomic norm minimization is proposed.Firstly the data fidelity between the low-bit quantized data and the unquantized data received at the array is established,and then the atomic norm minimization is applied to recover the DOAs.Thus the gridless DOA estimation problem can be transformed into the joint minimization of the mismatch between the quantized and unquantized measurements and the corresponding atomic norm defined by the received data.To efficiently solve the optimization problem,the accelerated proximal gradient technique is utilized,such that the solution can be iteratively updated by performing gradient descent with momentum and proximal projection.Compared with the classical subspace-based methods,the algorithm proposed in this paper is also applicable in the case of insufficient snapshots.Simulation results demonstrate the superiority of the proposed method both in accuracy and efficiency.

关 键 词：无网格化波达角估计 低比特量化 原子范数最小化 加速近端梯度 

分 类 号：TN957.51[电子电信—信号与信息处理]