The summational projection algorithm for the adaptive Volterra filter

In this paper, we propose a summational projection algorithm with block length control. This algorithm has the convergence properties of high speed and high accuracy under high noise for adaptive Volterra filters. And this algorithm has a computational complexity of O(p·N²). The proposed algorithm realizes these convergence properties by controlling the block length in the updating algorithm. In addition, the algorithm can track the variation of the impulse response of an unknown system and the power variation of an additive noise. We show the effectiveness of the proposed algorithm by computer simulations in this paper.