An algorithm for accelerating GHM Multiwavelet transformation on FPGA

Multiwavelets offer the possibility of superior performance for signal processing applications, compared with scalar wavelets. A multiwavelet system can simultaneously provide perfect reconstruction while preserving length, good performance at the boundaries, and a high order of approximation. The GHM Multiwavelet system is one of the most common multiwavelet systems with two pairs of scaling and wavelet functions. Hardware Acceleration of computation-intensive systems like GHM Multiwavelet will speed up such complicated routines. The ability to parallelize algorithms also makes the hardware accelerator more preferable. In this paper an attempt is made at obtaining a GHM Multiwavelet filter bank Hardware Accelerator for processing 1-Dimensional signals on FPGA.