论文部分内容阅读
针对欠定线性瞬时混合中的混合矩阵条件数较大的情况,使用最小二乘估计算法求解源信号造成的噪声放大问题,研究了一种求解分离矩阵的算法。此方法基于最大信干噪比的原则建立代价函数,通过最优化此函数得到分离矩阵。仿真实验表明,该算法与最小二乘算法相比可以改善在混合矩阵条件数较大时噪声放大影响下的分离性能。
Aiming at solving the problem of noise amplification caused by the source signal using the least-squares estimation algorithm, aiming at the condition that the number of mixing matrix in underdetermined linear instantaneous mixing is large, an algorithm to solve the separation matrix is proposed. This method establishes the cost function based on the principle of maximum SINR, and obtains the separation matrix by optimizing this function. Simulation results show that the proposed algorithm can improve the separation performance under the influence of noise amplification when the number of mixing matrix is larger than least squares algorithm.