电子学报 ›› 2018, Vol. 46 ›› Issue (6): 1384-1389.DOI: 10.3969/j.issn.0372-2112.2018.06.016

• 学术论文 • 上一篇    下一篇

SαS分布噪声环境下高分辨率二维DOA估计算法及快速实现

陈媛1, 刘金磊1, 孙奇福1, 阳小龙1, 于尧2   

  1. 1. 北京科技大学计算机与通信工程学院, 北京 100083;
    2. 东北大学高级通信与网络实验室, 辽宁沈阳 110819
  • 收稿日期:2017-03-27 修回日期:2017-08-17 出版日期:2018-06-25 发布日期:2018-06-25
  • 作者简介:陈媛,女,1988年生于山东菏泽.现为北京科技大学计算机与通信工程学院讲师.主要研究方向统计信号处理、脉冲噪声参数估计等.E-mail:chenyuan15@ustb.edu.cn;刘金磊,男,1993年生于河南鹿邑.现为北京科技大学计算机与通信工程学院研究生.主要研究方向为信号处理、无线通信等.E-mail:841304963@qq.com;孙奇福,男,1983年生于北京.现为北京科技大学计算机与通信工程学院副教授.主要研究方向为网络编码与信息理论.E-mail:qfsun@ustb.edu.cn;阳小龙,男,1971年生于四川冷水.现为北京科技大学计算机与通信工程学院教授、博士生导师.主要研究方向为IP网络理论与技术、网络新业务与安全等.E-mail:yangxl@ustb.edu.cn;于尧,女,1982年生于辽宁沈阳.现为东北大学副教授、硕士生导师.主要研究方向为通信网络和机器学习.E-mail:yuyao@mail.neu.edu.cn
  • 基金资助:
    装备预研教育部联合基金(No.6141A02033307);中央基本科研业务费新教师基金项目(No.2302015FRF-TP-15-033A1);博士后基金项目(No.2016M590047)

A High Resolution Algorithm for 2D-DOA Estimation and Its Fast Implementation in the Presence of SαS Distributed Noise

CHEN Yuan1, LIU Jin-lei1, SUN Qi-fu1, YANG Xiang-long1, YU Yao2   

  1. 1. School of Computer & Communication Engineering, University of Science & Technology Beijing, Beijing 100083, China;
    2. Advanced Communication and Networking Lab, Northeastern University, Shenyang, Liaoning 110819, China
  • Received:2017-03-27 Revised:2017-08-17 Online:2018-06-25 Published:2018-06-25

摘要: 本文设计出一种针对脉冲噪声的二维鲁棒高分辨率波达方向(DOA,Direction of Arrival)估计算法,以解决雷达、声纳等无线通信领域中脉冲噪声环境下IAA(Iterative Adaptive Approach)无法准确估计出DOA的问题.该算法中,用最小p阶范数代替WLS(Weighted Least Squares)作为最优化求解的代价函数.此外,根据Toeplitz-Block-Toeplitz(TBT)矩阵性质和FFT简化计算过程,提出该算法的快速实现方法,提高算法的计算效率.该算法在对称α-稳定(SαS,Symmetric Alpha-Stable)分布噪声环境下建模,仿真结果表明:与CRCO-MUSIC(CoRrentropy based COrrelation-MUltiple Signal Classification)算法和MUSIC-FLOM(MUltiple Signal Classification-Fractional Lower-Order Moment)算法相比,二维lp-IAA算法可以在低信噪比、单快拍条件下有效分辨出相邻多目标信号;快速算法可以在保证高分辨率的前提下,算法平均运算时间降低至原来的约1/40.

关键词: 脉冲噪声, SαS噪声, 波达方向, 鲁棒算法, 高分辨率, Toeplitz-Block-Toeplitz矩阵, 最小p阶范数

Abstract: This paper presents a robust and high-resolution 2D DOA (Direction of Arrival) algorithm in the field of wireless communication such as radar and sonar,solving the problem that IAA (Iterative Adaptive Aproach) fails to provide reliable DOA estimation performance in impulsive noise environment.In the algorithm,we replace WLS (Weighted Least Squares) with least p order norm (lp-norm) as the cost function.Furthermore,to decrease the complexity of the proposed method,a fast implementation of 2D lp-IAA is also developed based on simplifying the calculation of inverse TBT (Toeplitz-Block-Toeplitz) matrix and fast Fourier transform.Simulation results indicate that,under the condition of low GSNR (Generalized Signal-to-Noise Ratio) SαS distributed noise with single snapshot,2D lp-IAA algorithm performs better in distinguishing adjacent multi-target signals than CRCO-MUSIC (CoRrentropy based COrrelation-MUltiple Signal Classification) and MUSIC-FLOM (MUltiple Signal Classification-Fractional Lower-Order Moment).Moreover,the average computation time of the fast method is reduced to nearly 1/40 of the lp-IAA algorithm.

Key words: impulsive noise, SαS distributed noise, DOA, robust algorithm, high resolution, Toeplitz-Block-Toeplitz, least p order norm

中图分类号: