Abstract: To solve the problem of mixing matrix estimation for underdetermined blind source separation (UBSS) when the number of sources is unknown, this paper proposed a novel mixing matrix estimation method based on average information entropy and cluster validity index (CVI). Firstly, the initial cluster center is selected by using fuzzy C-means (FCM) algorithm and the corresponding membership matrix is obtained, and then the number of clusters is obtained by using the joint decision of CVI and average information entropy index of membership matrix, then multiple cluster number estimation results can be obtained by using multiple CVIs. Then, according to the results of the number of multiple clusters estimation, the number of radiation sources is determined according to the principle of the subordination of the minority to the majority. The cluster center vectors obtained from the clustering operation of the estimated number of radiation sources are fused, that is the mixing matrix is estimated based on the degree of similarity of the cluster center vectors. When the source signal is not sufficiently sparse, the time-frequency single source detection processing can be combined with the proposed method to estimate the mixing matrix. The effectiveness of the proposed method is validated by experiments.

Key words: UBSS, estimation of the number of radiation sources, mixing matrix, FCM clustering, CVI