航空学报 > 2023, Vol. 44 Issue (22): 629226-629226   doi: 10.7527/S1000-6893.2023.29226

集成先验的低复杂度Nyström-MUSIC方法

刘秋雨1,2, 蒋彦雯1,2(), 范红旗1,2, 连红飞1,2   

  1. 1.国防科技大学 自动目标识别全国重点实验室,长沙  410073
    2.国防科技大学 电子科学学院,长沙  410073
  • 收稿日期:2023-06-27 修回日期:2023-07-19 接受日期:2023-09-13 出版日期:2023-11-01 发布日期:2023-09-21
  • 通讯作者: 蒋彦雯 E-mail:jiangyanwen0712@nudt.edu.cn
  • 基金资助:
    国家重点研发计划(2022YFB3902400)

Low-complexity Nyström-MUSIC method based on prior information

Qiuyu LIU1,2, Yanwen JIANG1,2(), Hongqi FAN1,2, Hongfei LIAN1,2   

  1. 1.National Key Laboratory of Automatic Target Recognition,National University of Defense Technology,Changsha  410073,China
    2.College of Electronic Science and Technology,National University of Defense Technology,Changsha  410073,China
  • Received:2023-06-27 Revised:2023-07-19 Accepted:2023-09-13 Online:2023-11-01 Published:2023-09-21
  • Contact: Yanwen JIANG E-mail:jiangyanwen0712@nudt.edu.cn
  • Supported by:
    National Key Research and Development Program(2022YFB3902400)

摘要:

传统多重信号分类(MUSIC)方法具有较高的角度分辨率和估计精度,但其子空间分解和谱峰搜索的计算复杂度较高、实时性较差,在阵列雷达系统的实际应用中受到了极大限制。为解决算力有限时阵列雷达角度估计性能受限的难题,提出了集成先验的Nyström-MUSIC方法,先后通过Nyström近似降低协方差矩阵维度和先验角度信息缩小谱峰搜索范围,分别减小了子空间分解的矩阵规模和谱峰搜索次数,有效降低了MUSIC方法的计算复杂度。仿真表明:集成先验的Nyström-MUSIC方法在具备与传统MUSIC方法相当的分辨性能和估计精度的条件下,将算法复杂度降低了8倍,运行时长减少80%以上,实现了对目标方位的快速高精度估计。

关键词: 角度估计, 低复杂度, 先验信息, Nystr?m近似, 多重信号分类(MUSIC)

Abstract:

The traditional Multiple Signal Classification (MUSIC) method has high angle resolution and estimation accuracy, but the high computation complexity and poor real-timeliness involved in using the method for subspace decomposition and spectral peak search greatly limit its practical application in array radar systems. To address the problem that the angle estimation performance of array radar is limited when the computing power is limited, a low-complexity Nyström-MUSIC method based on prior information is proposed. The Nyström approximation is successively exploited to decrease the dimension of the covariance matrix and the prior angle information to determine the spectral peak search in a small area. The matrix size of the subspace decomposition and the time of spectral search are reduced, and finally the computation complexity of MUSIC method is effectively reduced. Simulation results illustrate that compared with the traditional MUSIC method, the low-complexity Nyström-MUSIC method based on prior information reduces the computation complexity by 8 times and cuts the running time by more than 80% under the condition of comparable resolution performance and estimation accuracy. It is verified that the method proposed can realize fast and high-precision estimation of the target azimuth.

Key words: angle estimation, low complexity, prior information, Nystr?m approximation, multiple signal classification (MUSIC)

中图分类号: