Adaptive filtering based on subband decomposition

One important approach to reduce the computational complexity of adaptive filters is that based on subband decomposition, in which the input signals are represented in terms of a set of N near orthogonal signal components using an orthogonal transformation. This is because the representation allows processing schemes in which each of these orthogonal signal components are independently processed. In this paper we present FIR and IIR adaptive filter structures based on subband decomposition in which the input signals are splite in a set of decomposition in which the input signals are split in a set of orthogonal signal components. A reduced order adaptive filter is then inserted in each subband, whose coefficients are independently updated. The computer simulation results show that the both types of subband adaptive filter structures reduce computational complexity considerably, and have fairly good convergence property.