Adaptive blind signal separation using a risk-sensitive criterion

An adaptive blind signal separation algorithm is proposed using a risk-sensitive criterion framework. The proposed criterion varies the consideration weight of the adaptation quantity depending on errors in the estimates: the adaptation is accelerated when the estimation error is large, and unnecessary acceleration of the adaptation does not occur close to convergence. Hence, this method converges faster than conventional least squares methods. The effectiveness of the new algorithm is demonstrated by simulation. © 2001 IEEE.