基于E-PASTd的盲扩频码序列估计算法

Blind Estimation of PN Spreading Sequence Based on E-PASTd

针对直接序列扩频信号的扩频码盲估计问题,改进了特征分解的方法,提出了一种基于压缩投影近似子空间跟踪的PN码恢复算法。该算法由两个主特征向量估计非同步延迟值,结合利用压缩投影近似子空间跟踪技术的快速收敛特性提取主分量,避免了对自相关矩阵的直接特征分解运算。计算机仿真表明该算法降低了数据的存储量,易于硬件实现,具有良好的收敛特性,性能优于已有的梯度算法和神经网络算法。

Aim at the problem of blind estimation of DS/SS spreading sequence, the approach of eigenvalue decomposition is modified, moreover a Pseudo Noise （PN） code recovery algorithm based on Projection Approximation Subspace Tracking with deflation （PASTd） is proposed. This algorithm estimates the desynchronized delay value through the principal eigenvectors, and extracts the principal component by using fast convergence properties of PASTd technique, avoiding direct eigenvalue decomposition. Computer simulations show that this algorithm decreases the storage consuming, is prone to hardware realization and has fast convergence properties. The performances are better than the gradient-descent algorithm and the neural networks algorithm.