A weighted overlap-add based wave domain adaptive filtering algorithm

The application of adaptive algorithms like noise reduction or echo cancellation on Wave Field Synthesis and Wave Field Analysis may be performed by using Wave Domain Adaptive Filtering in order to limit the computational complexity. The computational cost can be furthermore reduced taking advantage of filter banks theory. Moreover, the signal decomposition into subbands allows to achieve a faster convergence rate and a lower mean-square error for these adaptive algorithms. In this paper we introduce a Weigthed Overlap-Add algorithm for subband processing: it is used to design DFT filter banks where no relation between number of subbands and decimation factor is required. We will show how this scheme leads to the expected results and even to a better behaviour of the transformations involved in Wave Domain Adaptive Filtering.