基于GAMP算法的FDA-MIMO雷达的联合距离角度估计

Joint Range-Doa estimation based on GAMP for FDA-MIMO radar

MIMO雷达存在无法同时估计角度、距离的缺点,引入频率偏差带来距离维度信息(频控阵MIMO雷达),耦合了距离、角度信息,如何同时估计距离和角度成为提升系统能力的关键点。文章针对频控阵MIMO雷达系统中的角度、距离估计问题,提出了一种基于消息传递算法的近线性最小均匀误差估计算法。该算法根据角度分辨率和距离分辨率,将角度搜索空间和距离搜索空间划分为多个子区间,映射来波角度、距离门为稀疏向量。算法将角度估计、距离估计问题重构为稀疏线性模型,建立因子图,采用和积算法估计了该稀疏向量。同时,基于中心极限定理、大数定理近似和积算法中的消息,减少了计算复杂度。仿真结果表明,文章提出的算法可近似实现线性最小均匀误差估计,具有复杂度低、性能损失少、计算效率高等优点。

MIMO radar has the disadvantage of not being able to estimate both angle and distance simultaneously.The introduction of frequency deviation brings distance dimension information(FDA MIMO radar),which couples distance and angle information.How to estimate both distance and angle simultaneously has become a key point in improving system capabilities.In this paper,we proposed a low complexity and efficient near linear minimum uniform error estimation algorithm based on message passing algorithm for angle estimation in frequency diverse array MIMO(FDA-MIMO)radar systems.This algorithm divides the range of angle into multiple sub intervals based on angle resolution,maps the incoming direction of arrival as a sparse vector,reconstructs the angle estimation problem into a sparse linear model and establishes a factor graph,and uses the sum product method to estimate the sparse vector.At the same time,based on the central limit theorem,and the large number theorem approximation,the computational complexity in the product method is reduced.The simulation results show that the algorithm proposed in this paper can approximately achieve LMMSE estimation,but its complexity is low.