一种分数低阶时频分解的盲波束形成算法  

Blind beamforming algorithm based on fractional lower-order time-frequency decomposition

作　　者：陈沛 赵拥军[1] 刘成城[1] 李海文 CHEN Pei ZHAO Yongjun LIU Chengcheng LI Haiwen(Institute of Navigation and Aerospace Target Engineering, Information Engineering Univ., Zhengzhou 450001, China Chongqing Communication College, Chongqing 400035, China)

机构地区：[1]信息工程大学导航与空天目标工程学院,河南郑州450001 [2]重庆通信学院,重庆400035

出　　处：《西安电子科技大学学报》2017年第5期134-139,共6页Journal of Xidian University

基　　金：国家自然科学基金资助项目(61401469)

摘　　要：针对现有盲波束形成算法在脉冲噪声环境中性能的下降,提出一种基于分数低阶时频分解的盲波束形成算法.该算法首先将分数低阶统计量引入传统的短时傅里叶变换中,实现抑制脉冲噪声的时频分解.然后利用对各阵元的接收信号进行分数低阶时频分解的结果,结合聚类和不确定集方法,实现导向矢量的最优估计.最后利用最小方差无畸变响应算法获得最优权矢量,实现了高效的多目标盲波束形成.仿真实验结果表明,该算法在服从对称alpha稳定分布的脉冲噪声环境中,较现有盲波束形成算法的输出性能更优.该算法不依赖源信号自身特性,适用范围更广.A novel blind beamforming algorithm based on fractional lower-order time-frequency decomposition （TFD） is proposed to improve the performance of existing blind beamforming algorithms in impulsive noise environment. The traditional short-time Fourier transform （STFT） is first improved by utilizing fractional lower-order statistics to realize TFD in impulsive noise environment. Then, by combining the clustering method and the method for the uncertainty set, the TFD results of the receiving data at each sensor are used to realize the estimation of steering vectors （SV）. Finally, the optimal weight coefficients of the proposed blind beamformer are achieved by substituting the estimated SV into the minimum variance distortionless response （MVDR） beamformer. Simulation results demonstrate that this algorithm can achieve superior output performance over the existing blind beamforming methods in impulsive noise environment. The proposed algorithm hardly needs any specific property of the receiving signals and can be more widely used.

关 键 词：时频分解 分数低阶统计量 盲波束形成 对称alpha稳定分布 

分 类 号：TN911.7[电子电信—通信与信息系统]