Abstract of my Final Project

Here is my final project abstract. The title of my final project is Analysis of Convolutional Coding with Viterbi Decoding Performance via Frequency Non-Selective Rayleigh Fading. You know, this topic came to me when I went to Electrical Engineering Department Library. Mmm, I was inspired by a final project by Yudi Darmadi, EL 99. Thanks, Mas… 🙂

Abstract:

One of the channel coding deployment methods is utilizing convolutional encoder in transmitter side and Viterbi decoder in receiver side. Some parameters influence their performance in correcting error. They are constraint length, polynomial generator, bit number of soft decision, and code rate. This final project gives details about varying constraint length and bit number of soft decision that affect the system performance which is indicated by Bit Error Rate (BER) and energy per normal noise E_b/N_0­­ correlation. It also gives details the effects of interleaver and deinterleaver attachment to the communication system performance.

We used computer simulation for modelling the system. There is a final project which explained this problem but only applied Additive White Gaussian Noise (AWGN) channel in the system. Because AWGN channel is not practical, we use frequency non-selective Rayleigh fading channel as a channel transmission.

Here are the results of the simulations. To reach BER 10^-3­, at 3-bit soft decision, without interleaver, and 3, 5, 7, 9 constraint length, we need E_b/N₀ at 17, 14, 15, and 12.5 dB. It shows that bigger constraint length, lesser the requirements of E_b/N₀. Moreover, utilization of 3-bit soft decision gives 2-dB performance better than hard decision. However, more number of bit soft decision, for 10-bit soft decision, the performance is unchanging, but the delay is increase and complexity as well. When we applied interleaver and deinterleaver for 10 bit in the system, to reach BER 10^-3, at 3-bit soft decision and constraint length 9, we need E_b/N₀ at 6 dB.

Keywords:

Viterbi decoder, Rayleigh fading channel, soft decision, constraint length, generator polynomial, code rate, convolutional encoder