一种DS/SS信号PN码序列估计的神经网络方法

A Neural Netword Approach to Estimation of PN Spreading Sequence in DS/SS Signals

本文提出了一种利用神经网络（Ｎ．Ｎ．）完成直接序列扩频（ＤＳ／ＳＳ）信号 ＰＮ（伪噪声码）码序列估计的新方法。该方法充分利用了无监督Ｎ．Ｎ．的自适应主元提取特性，加上其高速运算能力，解决了传统方法在ＰＮ码的实时处理及实现上的困难，理论分析和数值结果表明了本方案能在较低的输入信噪比条件下完成对ＰＮ码的正确估计。

In this paper. We propose an approach that can estimate the PN sequence in the lower S/N ratio DS/SS signals, this method utilize the characteristics of the self-organization and the principle component extraction of the unsupervised neural networks adequately, in additional to its higher-speed operaion ability, successfully solve the conventional questions about PN sequence. The theoretic analysis and experimental results show that the approach can work well on the lower S/N ratio input signals.