A Proposed SM-IMSAF Algorithm with Fast Convergence Rate

In order to obtain a fast convergence rate, we propose a novel algorithm at the basis of the improved multiband-structured subband adaptive filter algorithm (IMSAF). The proposed algorithm incorporates the idea of set-membership into the IMSAF (SM-IMSAF). The update equation of the proposed SM-IMSAF is derived by using the Lagrange Multiplier method. Due to the effect of set-membership, the proposed SM-IMSAF achieves a better performance than some existing well-known algorithms. The simulation experiments are carried out under the condition of the system identification applications. Considering the practical condition, exact-modeling as well as under-modeling is taken into account in the simulations. At the same time, the tracking ability of SM-IMSAF algorithm is also researched when the unknown system mutates. The simulation results verify the superiority of the SM-IMSAF algorithm.