Underdetermined blind source separation method applying multiple constraints

Abstract

The invention discloses an underdetermined blind source separation method applying multiple constraints, which is improved on the basis of a traditional blind source separation method to realize a better signal separating effect. The method disclosed by the invention aims to separate mixed signals and comprises the steps of firstly carrying out equalization and whitening to improve the robustness of initial conditions, carrying out multi-constraint limit on a non-negative matrix factorization algorithm and carrying out optimization on an objective function and further improving the signal separating property through a feedback mechanism finally. Therefore, the method disclosed by the invention has the advantages of good factor interpretability and high separated signal purity.

Description

technical field [0001] The invention relates to an underdetermined blind source separation method imposing multiple constraints, belonging to the technical field of signal processing. Background technique [0002] BSS (Blind Source Separation, Blind Source Separation) refers to the technology of recovering the source signal only by the signal received by the sensor when the source signal and the mixing method are unknown. Blind source separation when the number of sensors is less than the number of source signals, called underdetermined blind source separation, is a hot research issue in the field of blind signal processing in recent years. All have broad application prospects. [0003] Among the traditional blind signal separation methods, the ICA algorithm is developed on the basis of the principal component analysis (PCA) method, which is only suitable for overdetermined or positive definite mixed models, and must have certain assumptions and constraints. Nonnegative ma...

AI Technical Summary

Problems solved by technology

[0004] At present, most of the research is to use the sparsity of the signal to solve the problem of underdetermined blind separation based on the sparse component analysis algorithm. M. Zibulevsky et al. use the sparse component analysis to propose a two-step method to estimate the mixing matrix and the source signal, the mixing matrix The quality of the estimation directly affects the subsequent signal separation effect; based on the sparse representation, Bofill uses the clustering method and the shortest path method to estimate the mixing matrix and the source signal respectively, and successfully separates six source signals from the two observed mixed signals, but The sparsity requirement of the signal is strong; in view of the fact that the basic NMF algorithm cannot solve the problem of blind source separation under underdetermined conditions, Cichocki et al. proposed a multi-layer NMF algorithm to achieve extremely sparse blind separation for each layer of decomposed signals, but The implementation of the algorithm is relatively complicated; Huang Yuhan used the non-negative matrix factorization algorithm (IS-NMF) based on Itakura-Saito divergence to study the separation of single-channel music signals, and the signal sparsity after time-frequency domain transformation is still not ideal , which leads to a large estimation error of the mixing matrix and affects the performance of the entire algorithm. Therefore, reducing the sparsity requirements of the source signal and algorithm complexity and improving the separation accuracy are research directions with theoretical significance and economic value.

Images