基于神经网络的方法获取DS／SS信号的PN码

提出了两种基于神经网络(NN)获取直接序列扩频(DS/SS)信号的伪随机码(Pseudo RandomCode)的方法。这两种方法充分利用了神经网络无监督的自适应学习方法来提取DS/SS信号的协方差矩阵主特征向量,并可以进行并行运算,因此有较快的运算速度。同时文章对这两种算法进行了性能比较,试验结果和理论分析表明,这两种方法可以在较低的输入信噪比的条件下较好地估计DS/SS信号的PN码。

PN Code Acquisition of DS/SS Signals Based on Neural Network Methods

Two methods for acquiring the PN code of DS/SS signals based on Neural Networks are advanced in this article. Both methods use unsupervised adaptive learning of neural networks to compute the principal eigen-vector of the covariance matrix of DS/SS signal with the parallel computation methods, and can get a faster computation speed. A comparison of the two methods is also included in this article. The test result and the theory analysis show that both methods can acquire the PN code of DS/SS signals in a very low input SNR environment.