Abstract
Recently, the data-selective adaptive Volterra filters have been proposed;however, up to now, there are not any theoretical analyses on its behaviorrather than numerical simulations. Therefore, in this paper, we analyze therobustness (in the sense of l2-stability) of the data-selective Volterranormalized least-mean-square (DS-VNLMS) algorithm. First, we study the localrobustness of this algorithm at any iteration, then we propose a global boundfor the error/discrepancy in the coefficient vector. Also, we demonstrate thatthe DS-VNLMS algorithm improves the parameter estimation for the majority ofthe iterations that an update is implemented. Moreover, we prove that if thenoise bound is known, we can set the DS-VNLMS so that it never degrades theestimate. The simulation results corroborate the validity of the executedanalysis and demonstrate that the DS-VNLMS algorithm is robust against noise,no matter how its parameters are adopted.