A Dynamic Step-Size Updating Rule for Single Distribution Blind Source Separation

Abstract

A novel dynamic step-size updating rule for online blind separation of single distribution source is presented in this paper. The rule dynamically divides the process of separation into two stages by using the independent measurement of the output signals, and respectively adopts corresponding step-size updating algorithm. The rule can duly detect whether the channel matrix is changed or not, so it can separate source signals effectively under the situation for variable channel or invariable channel. The simulations have verified that this rule is robust and retains not only the faster convergence rate but also the lower steady-state error.