Fast separation algorithm of mixed signals based on GEVD

In order to improve the efficiency of mixed signal separation, we propose a fast algorithm based on generalized eigenvalue decomposition. On the basis of removing redundant signals, the observation matrix bundle is established by FIR filtering method, and then the source signal is separated by solving its eigenvector and eigenvalue matrix. Theoretical analysis and simulation experiments show that the proposed algorithm can effectively separate source signals, and the separation efficiency is greatly improved compared with existing algorithms.