中文摘要：

本文提出一种基于广义能量函数（GEF）的直接序列扩频（DS／SS）信号扩频码序列的盲估计方法。广义能量函数通过引入一个加权矩阵来优化线性神经网络的连接权矢量，可以推导出一种新的递归最小二乘（RLS）学习算法。该算法能高效提取一个输入信号相关矩阵的多个主分量，可对同步和非同步模型下的PN码序列实现盲估计。该算法具有收敛快、稳健性好等优点，其运算量和储存量远远小于特征值分解算法，收敛速度、相关性能和运算复杂度等恢复性能优于压缩投影逼近子空间跟踪（PASTd）算法和改进神经网络（MHR）算法。计算机仿真证明，该算法能在较低的信噪比条件下，实时高效地恢复PN码序列，具有优异的性能。

英文摘要：

This paper provides a Generalized Energy Function （GEF） to search for the optimum weights to estimate the PN spreading sequence by introducing a weighting matrix, which can parallel and real-time acquire the principal eigenvectors of the covariance with high efficiency. The GEF algorithm can work by real-time and be easily implemented both in synchronization and asynchronization signal modes. Computer simulations show that this algorithm can estimate PN spreading sequence quickly and accurately at low Signal-to-Noise Ratio （SNR）, largely reducing computational complexity and storage than EVD. Furthermore,its properties of convergence, correlation and computational storage are better than Projection Approximation Subspace Tracking （PAST） and Modified Hebbian Rule （MHR） algorithms.