An important problem in adaptive algorithm is opting step size.This paper establishes a nonlinear functional relationship between step size and measure of signal of dependence.On the basis of the functional relationship

the paper presents the new algorithm of variable step size adaptive ICA and analyses convergence and steadiness of the algorithm.This algorithm of variable step size on the basis of the degree of signal separation which is reflected by measure of signal dependence overcomes the disadvantages of traditional algorithms in the process of step size change of adaptive steady state.The theoretical analysis and computer simulation results show this algorithm has better performance in convergence and steadiness.