Implementation of the Standard Floating Point DWT Using IEEE 754 Floating Point MAC

Abstract

This work concentrates mainly for the implementation of Standard DWT using IEEE 754 floating point format. Currently, in the signal processing, for audio purpose the fixed point DWT is used as audio CODEC [1]. The main bottleneck of the fixed point DWT or the traditional DWT is the speed because at the input of the fixed point DWT the over-sampled ADC which is the Sigma-Delta ADC is used. The Sigma-Delta ADC can’t give the speed more than 1 MHz because as the sampling rate increases, the step size decreases so that it takes more time to follow the analog signal which causes the limitation of the speed. Due to the speed limitation of ADC, the whole audio CODEC system which was designed by the fixed point DWT becomes slow even it has the capability to operate with a better speed. Hence, to optimize the system the FIR filters which are used to constitute the standard floating point DWT have been implemented in VLSI.