DSP implementation of Viterbi decoder

OData support
Supervisor:
Dr. Gaál József
Department of Networked Systems and Services

The subject of my thesis is an implementation of software Viterbi decoder on a multi core, generic Texas Instruments DSP and analyzing the properties of the implemented decoder.

During my task I studied the concept of convolutional coding, and Viterbi algorithm as optimal convolutional decoding algorithm. I analyzed the theoretical expectations of convolutional codes with Matlab simulation software. For this, I created a simulation enviroment with Matlab Simulink. With the help of the Simulink models I did Bit Error Rate analysis, with different convolutional code parameters and different modulation techniques.

After Matlab simulations, I implemented a software Viterbi decoder with C language, running on Texas Instruments TMS320C6678 DSP. For measuring the decoding propreties of the implemented Viterbi decoder, I developed measuring and simulation enviroment. In creating the simulation enviroment, I followed the concepts of the Matlab models and implemented those in C language. I have tested the Viterbi decoder and the simulation enviroment on the TMDXEVM6678 evaluation board, and I did multiple measuments using this board.

I analyzed the results of measurments and evaluated the usage possibilites and limits of the implemented software Viterbi decoder.

Downloads

Please sign in to download the files of this thesis.