论文部分内容阅读
针对欠定盲源分离中混合矩阵的估计问题,以及现有算法大多存在复杂度高、估计精度低的缺陷,在分析K-Plane算法的基础上,提出了一种改进的欠定混合矩阵估计算法——IK-Plane(improved K-Plane)算法.IK-Plane算法通过最优化方法,计算与所有观测信号的内积和最小的向量,并将该向量作为新的法向量,改进了法向量的更新方法,从而改善了算法的时间复杂度及估计精度.实验结果表明:相对于K-Plane算法,IKPlane算法在提高估计精度的同时,能够显著地降低算法的时间复杂度.
In order to solve the problem of hybrid matrix estimation in underdetermined blind source separation and most existing algorithms have the defects of high complexity and low estimation accuracy, based on the analysis of K-Plane algorithm, an improved underdetermined mixed matrix estimation Algorithm - IK-Plane (improved K-Plane) algorithm. The IK-Plane algorithm uses the optimization method to calculate the inner product and the smallest vector with all the observed signals, and uses the vector as a new normal vector to improve the normal vector , Which improves the time complexity and accuracy of the algorithm.The experimental results show that compared with the K-Plane algorithm, the IKPlane algorithm can significantly reduce the time complexity of the algorithm while improving the estimation accuracy.