基于Givens旋转多用户多径DS-CDMA信号伪码盲估计

针对多径多用户低信噪比下的伪码估计,提出一种基于Givens旋转多用户多径DS-CDMA信号伪码盲估计的算法.该算法在同步接收扩频信号的基础上,利用特征值分解法获得用户特征向量子空间,通过2维Givens矩阵对子空间特征向量组成的2维载荷矩阵进行旋转,当定义的载荷矩阵总方差达到极小值时停止旋转,得到伪码序列的最佳估计.该算法克服了传统特征分解法估计多用户伪码序列时的缺陷,解决了多用户多径低信噪比情况下伪码序列的盲估计问题.仿真结果表明该算法是有效的,且具有抗多径干扰的能力.

Multi-user multi-path DS-CDMA signal pseudo-code blind estimation based on Givens rotation

For the pseudo-code estimation problem of multipath coherence multiuser(MCM) DS-CDMA(direct sequence-code division multiple access) system in low SNR(signal to noise ratio),a new pseudo-code blind estimation method by decomposing the eigenvectors that were based on the Givens rotation was proposed in this paper.The method of eigenvalue decomposition was used to get the users eigenvectors subspace,under the condition of receiving the signal synchronously; and then the two-dimensional loading matrix that consisted of subspace eigenvectors was rotated with the two dimensional Givens rotation matrix until the total variance function of the loading matrix defined in the article reached the smallest value,at which time the estimation of the pseudo-code sequences was the best one.The method could overcome the defects of traditional decomposing eigenvectors for multi-user pseudo-codes' estimation,and it was good at anti-multipath interference.It could realize blind estimation of the multi-user multi-path pseudo-code sequences in low SNR.The simulation results showed that the effectiveness and the anti-multipath interference capability of this method.