相干多径环境下信号空间特征及波达方向估计  

Estimation of spatial signature and direction of arrivals for signals in coherent multipath environment

作　　者：周围[1] 朱联祥[2] 周正中[1] 张德民[2] 

机构地区：[1]电子科技大学通信与信息工程学院,四川成都610054 [2]重庆邮电大学移动通信技术重点实验室,重庆400065

出　　处：《电波科学学报》2007年第4期685-690,共6页Chinese Journal of Radio Science

基　　金：国家自然科学基金(#60572089);重庆邮电大学4G移动通信专项资金(A2004-70);重庆邮电大学青年教师基金(A2005-30)

摘　　要：以均匀直线阵和四阶累积量为基础,提出了一种新的空间特征盲估计算法。该算法首先利用阵列输出四阶累积量构造了一种空间特征矩阵,对其作特征分解可以得到各用户的空间特征估计。然后,在空间特征估计基础上,结合前向/共轭后向空间平滑技术进行了多径波达方向(Direction of Arrival,DOA)估计。该方法突破了传统MUSIC、ESPRIT等算法的局限,可以利用M阵元估计2M2/3个DOA,且各DOA与信源自动配对。该算法不依赖于信号具体特征,适用于任意加性高斯噪声(如有色噪声)环境。理论分析和仿真结果说明了算法的有效性和鲁棒性。Based on uniform linear array and fourth order cumulant, a novel blind algorithm for spatial signature estimation was presented. Firstly, using the fourth order cumulant of array output, the algorithm constructed a spatial signature matrix, which can be used to estimate spatial signature via eigen decomposition. Then, based on spatial signature estimation and followed by forward/conjugatebackward-spatial-smoothing-technique, the algorithm was used to estimate the direction-of-arrivals （DOAs） of multipaths. The algorithm breaks through the limit of MUSIC and ESPRIT, can estimate 2M^2/a DOAs through M elements, and each DOA can automatically match the sources. The algorithm is not depended upon the specific characteristic of signals, and can be employed to additive arbitrary Gaussian noise （such as colored noise）. Theoretical analysis and computer simulation confirmed the availability and robustness of the algorithm.

关 键 词：智能天线 空间特征 波达方向(DOA) 高阶累积量 均匀直线阵 

分 类 号：TN929.5[电子电信—通信与信息系统]