FPGA implementation of 4 samples DWT based on the model of pyramidal structural data coding

In this paper, fast algorithm for discrete wavelet transform is analyzed, a signal processing model based on pyramidal structural data coding is described, and an implementation architecture of 4 samples DWT for both decomposition and reconstruction is proposed, in which we put only one filter into use. With the multiplier reuse method and techniques such as bit-moving, basic operation, by the language VHDL on the platform of MAXPLUSII, FPGA simulation and implementation for this architecture are fulfilled. Numerical experiments show that our result can decompose a four-sample discrete signal sequence and then reconstruct it successfully.