Separability Theory for Blind Signal Separation
盲分离问题的可分性理论

The basic theory of blind signal separation is discussed. Not only the sufficiency-neces-sity condition is given for the extraction of single source but also the problem how many signals can be extracted is answered. Moreover, the blind separation algorithm based on this theory is proposed. The method can also be applied to blind extraction in the case of morbid mixture. At last, simulation is done in different mixed conditions, and the results show that the separation theory is correct and the algorithm is valid.