空间特征的累量域估计算法及其在智能上行多波束形成中的应用

Spatial Signature Estimation Algorithm in Cumulant Domain and Its Application in Smart Uplink Multi Beamforming

在智能天线系统中，空间特征估计是实现空分多址的关键问题之一。本文提出了一种基于累量的多用户空间特征估计算法，并将该算法用于智能天线中智能上行的实现。该算法既不依赖于信号的的具体结构也不依赖于信道的模型假设，构造了累量域空间特征矩阵，通过对空间特征矩阵的特征分解得到各用户信号的空间特征估计。并以此为基础，设计了用于上行多波束成形的空间滤波器组，实现了上行多用户信号的分离。仿真实验表明，空间滤波器的响应可用于多用户相干源的波达方向估计。

In smart antenna system (SAS), spatial signature estimation is one of the key issues in implementing space division multiple access (SDMA). In this paper, a subspace based multiuser spatial signature estimation algorithm is presented and applied to implementing SDMA in SAS. The algorithm depends on neither the definite signal structure nor the assumption of channel model, it constructs a new matrix called spatial signature matrix (SS matrix) in cumulant domain, and estimates the multiuser spatial signature through eigen decomposition of SS matrix. Based on the spatial signature estimation, an uplink spatial filter bank is designed to implement the uplink multiuser signal separation. Computer simulation shows that the spatial response of spatial filter bank can be used to find DOA of co channel correlated multiuser signals.