多径信道中扩频信号伪随机序列盲估计

为解决异步直扩序列的盲估计问题，提出采用独立分量分析（ＩＣＡ）结合主分量分析（ＰＣＡ），在信道未知情况下有效盲分离多用户多径直扩信号和扩频码信息.由于不同用户信息是相互独立的且扩频序列是非相关的，对于某个用户，如果不存在多径效应，直接采用独立分量分析估计得到扩频序列；对于多径信号，进一步利用m序列的三阶相关特性估计出扩频序列.在高斯噪声下的性能分析表明，独立分量分析结合m序列三阶相关特性的盲估计过程，可以在不知道扩频码同步信息的情况下对异步多用户信号进行盲分离和估计.算法利用了ＣＤＭＡ系统的扩频增益，可以在低信噪比条件下盲估计扩频序列.

Blind estimation of pseudo-random sequences of spread spectrum signals in multi-paths

In order to blind estimation the sequence of asynchronous direct spread signals, this work used independent component analysis(ICA) combining with principle component analysis(PCA) to blindly separate the data bits of different users without the knowledge of the channel.Due to the independence of data bits of different users and the quasi-orthogonal of direct sequences,the direct sequence of signals without multi-path interference could be directly obtained by ICA,and the direct sequence of signals with multi-path interference could be further identified according to the triple correlation function(TCF) of m-sequence.This blind estimation method does not need the synchronization information of direct sequence.The performance analysis under Gaussian noise indicates that this blind estimation arithmetic utilizes the spread gain of CDMA systems and the m-sequences can be obtained in lower SNR signal.