| 摘要: |
| 针对稀疏阵列下的DOA估计中原子范数最小化的交替方向乘子算法存在精度偏低、运算量大和抗噪声性能较差的问题，提出一种基于MM优化的改进原子范数最小化算法（ANM-ADMM-MM）。该方法首先进行采样和量化，建立一个观测信号模型，通过原子范数最小化，得到半正定规划（SDP）形式；然后在ADMM算法交替迭代最小化代价函数中加入MM优化得到最优估计的Toeplitz矩阵；最后利用MUSIC算法对Toeplitz矩阵峰值搜索求得角度值。仿真验证表明，该方法在和ANM-ADMM算法相比，运行时间降低约74%，在信噪比为0 dB下的估计精度提高约55%，在信噪比为15 dB下的估计精度提高约35%。 |
| 关键词: 稀疏阵列 DOA估计 原子范数最小化 交替方向乘子算法 MM优化 |
| DOI：DOI:10.3969/j.issn.1672-2337.2026.02.008 |
| 分类号:TN957.51 |
| 基金项目:国家自然科学基金（62361015）；广西自然科学基金（2023GXNSFAA026060）；广西八桂学者专项经费资助 |
|
| Sparse Array DOA Estimation Method Based on ANM-ADMM-MM |
|
ZHANG Peihua, JIN Liangnian
|
|
1. Institute of Information and Communication, Guilin University of Electronic Technology, Guilin 541004, China;2. Key Laboratory of Wireless Broadband Communication and Signal Processing, Guilin 541004, China
|
| Abstract: |
| The alternating direction method of multipliers algorithm for atomic norm minimization in sparse array DOA estimation suffers from low accuracy, large amount of computation and poor noise robustness. An improved atomic norm minimization algorithm based on Majorization-Minimization （ANM-ADMM-MM） is proposed. Firstly, the method is sampled and quantized, and an observation signal model is established. The semi-definite programming （ SDP ） form is obtained by minimizing the atomic norm.Then, an MM optimization is incorporated into the ADMM iterative process to obtain the optimal estimated Toeplitz matrix. Finally, the MUSIC algorithm is employed to search for the peak of the Toeplitz matrix to determine the angles. Simulation results demonstrate that, compared with the ANM-ADMM algorithm, the running time of the proposed method is reduced by approximately 74%, the estimation accuracy is improved by about 55% at an SNR of 0 dB and by about 35% under the SNR of 15 dB. |
| Key words: sparse array DOA estimation atomic norm minimization alternating direction method of multipliers MM optimization |
