Normalized Combinations of Proportionate Affine Projection Sign Subband Adaptive Filter

Figure 14

NMSD curves of various algorithms with SNR = 30 dB, P_r = 0.001 for double-talk speech input.