Adaptive filters requiring zero multiplications

This paper introduces an adaptive filter structure that requires zero multiplications for its implementations. The primary input signals are quantized using DPCM and the DPCM outputs are processed by the adaptive filter. The sign algorithm. We show that if the parameters are chosen properly, hardware implementation of this filter structure requires no multipliers. Under the assumption that the signals are zero mean, wide-sense stationary, and Gaussian random processes, we derive theoretical results for the mean and mean-squared behavior of the filter. A simulation example is presented that shows very good match between theoretical and empirical results.