Estimation of Underdetermined Sources of Co Frequency Interference Signal Based on GAVMD-SVD

Sharing the same frequency band for electronic equipment is easy to cause mutual interference of the same frequency, and the effective way to reduce interference is blind source separation technology. One of the necessary conditions for its implementation is to accurately estimate the number of source signals. In order to solve the problem that the number of the same frequency source signals is difficult to estimate accurately in the underdetermined case, a method for estimating the number of underdetermined sources based on genetic algorithm optimization of the combination of variational mode decomposition (GAVMD) and singular value decomposition (SVD) is proposed. First, the number of modes and penalty factors are combined and optimized through genetic algorithm, then VMD decomposition is performed on the observed signal using the optimal combination parameters, and new modal functions are obtained by combining multiple groups of modal components, and then singular value decomposition is performed on its covariance matrix. Finally, the estimated value of the source number is obtained through the analysis of singular values. The simulation results show that the method has good adaptability and estimation effect.