Novel algorithm for underdetermined blind separation based on Sparse Component Analysis

The blind separation problem for sources that are sparse insufficiently is researched. The Sparse Component Analysis (SCA) algorithm is widely used to separate the linear mixtures when there are more sources than sensors. This paper presents a novel underdetermined blind source separation algorithm using sparse component analysis. The separation procedure has two steps: estimating mixing matrix and reconstructing source signals. We estimate the mixing matrix using clustering algorithm based on grid and density, and it can estimate mixing matrix better. When recovering source signals, a simpler method is used to get l₁ norm minimization solution. Simulation results showed that our method had a promising performance.