干扰子空间盲自适应估计的跳频多址信号检测

Multi-access Signal Detection for Frequency Hopping Based on Interference Subspace Estimation

针对跳频多址(FHMA)系统,提出了干扰子空间线性最小均方误差(MMSE)检测器和干扰子空间的快速跟踪算法。该算法通过添加噪声子空间得到了整个干扰自相关矩阵的权矢量,避免了同时对特征值对角阵和特征向量矩阵进行跟踪。结合神经网络学习机制中的NIC(NovelInformation Criterion)准则,进行权矢量的快速更新,降低了复杂度。仿真证明,该检测器具有很好的误码性能,收敛速度更快。

A Interference-space-Based blind adaptive Minimum Mean Square Error （MMSE） liner detector and faster subspace tracking algorithm were proposed in Frequency Hopping Multiple Access （FHMA） system. We got the optimum weights based whole interference coorelation matrix by adding noise-space and it needn＇t tracking the eigenvalue matrix and eigenvector matrix in this algorithm. The optimum weights was updated consulting Novel Information Criterion method in the neural network, the complexity was reduced greatly. The simulation result indicates that the detector has a good performance at bit-error-rate and faster speed in constringency.