Multidimensional DFT-modulated filter bank: Design and implementation
Maxim Vashkevich, Alexander Alexandrovich Petrovsky · 2014
The paper presents an approach to subchannel decomposition of multidimensional signals that is performed by means of a DFT-modulated filter bank designed for multidimensional signals. An r-dimension signal can be decomposed using r scalar DFT-modulated filter banks. However, the proposed vectorized filter bank allows to obtain polyphase filter bank structure that can be efficiently implemented in hardware or in software using vector processors.