基于改进ICA-R算法的多用户信号盲提取

Blind Extraction of Multi-User Signals based on Modified ICA-R Algorithm

针对单通道通信系统中多用户信号盲源分离(Blind Source Separation,BSS)时出现的干扰问题,提出了一种改进的带参考向量的独立成分分析(Independent Component Analysis with Reference,ICA-R)算法。该算法通过将接近性度量函数的倒数添加到对比函数中,从而得到一个新的对比函数,然后利用拉格朗日乘子法得到最优的权向量,最后,通过线性转换提取出多用户信号。提出的改进的参考独立分量分析算法与先进的ICA-R方法相比具有更快的收敛速度和更高的提取精度。仿真结果表明,该算法能够有效地提取出多用户信号,并且提取精度高,算法的鲁棒性好。

Aiming at the interference of muhi-user signals BSS （Blind Source Separation） in single-channel communication system, a modified ICA-R （Independent Component Analysis with Reference） algorithm with reference vector is proposed. By adding the reciprocal of similarity measure to the contrast function, a novel contrast function is derived, and then the optimal weighted vector acquired with Lagrange multiplier method, and finally multi-user signals extracted via a special linear transformation. Compared with state-of-the-art ICA-R methods, the proposed modified ICA-R algorithm enjoys a faster convergence speed and higher extraction quality. Simulation results indicate that the proposed algorithm could effectively extract the multi-user signals, and is of fairly high extraction precision and robustness.