Performance Analysis of Improved Sub-band Sparse Adaptive Algorithm for Acoustic Channel Identification

In this paper, we present the performance analysis of new Sub-band Improved Proportionate NLMS algorithm to improve the convergence rate of dispersive acoustical channel identification. Although the existing proportionate algorithms lately implemented had a good convergence speed in sparse systems, however they suffer from slow convergence speed in case of dispersive channels. This proposed algorithm based on Improved Proportionate NLMS incorporated into sub-band adaptive filtering structure denoted by S-IPNLMS shows good results in the latter case. In this approach, sub-band adaptive filtering is utilized to address challenges encountered in long adaptive filters, such as their slow convergence rate. Simulation results show that the sub-band IPNLMS performs better than the sub-band NLMS, PNLMS when the acoustical impulse response is dispersive.