Date of Award

1998

Degree Type

Thesis

Degree Name

Bachelor of Engineering Honours

Faculty

Faculty of Communications, Health and Science

First Advisor

Dr Stefan Lachowicz

Abstract

This thesis discusses the findings of the final year project involving Gallium Arsenide implementation of a triangular FIR filter to perform discrete wavelet transforms. The overall characteristics of Gallium Arsenide technology- its construction, behaviour and electrical charactersitics as they apply to VLSI technology - were investigated in this project. In depth understanding of its architecture is required to be able to understand the various design techniques employed. A comparison of Silicon and GaAs performance and other characteristics has also been made to fully justify the choice of this material for system implementation. A lot of research and active interest has gone into the field of image and video compression. Wavelet-based image transformation is one of the very efficient compression techniques used. An analysis of discrete wavelet transformations and the required triangular FIR filter was done to be able to produce a transform algorithm and the related filter architecture. Finally, the filter architecture was implemented as a VLSI design and layout. A variety of functional blocks required for the architecture were designed, tested and analysed. All these blocks were integrated to produce a model of a complete filter cell. The filter implementation was designed to be self-timed - without a system clock. Self-timed systems have considerable advantages over clocked architectures. Various design styles and handshaking mechanisms involved in designing a self-timed system were analysed and designed. There are many avenues still to explore. One of them is the VHDL analysis of filter architecture. Further development on this project would involve integration of higher-level logic and formation of a complete filter array.

Share

 
COinS