经典雷达测角方法采用解卷积技术引入病态性,通常具有较差的抗噪性能。文中首先分析解卷积病态性产生原因,并从信号建模出发,将测角过程看作是求解线性方程问题,并引入范数正则化技术,提出了基于TSVD及Tikhonov正则化方法的测角方法,可以较好地解决病态性问题。最后提出一种稀疏先验信息约束下的OMP测角方法。仿真结果表明,相比经典解卷积测角技术,所提方法均可取得更优的测角性能。

Conventional angle measurement methods are based on deconvolution algorithms, and usually have poor performances in low SNR situation because of introducing ill-condition. This paper firstly analyzes the reason of ill-condition in deconvolution, then based on the signal model, angle measurement process can be treated as solving linear equation problem. Further the norm regularization techniques are utilized in this paper for solving the problem. Finally, we develop a general method based on TSVD and Tikhonov regularizations, and a sparse method based on Orthogonal Matching Pursuit (OMP) under the condition of sparse prior constraint. Simulation results show that the proposed methods outperform conventional methods in angle resolution performances.