Study Of The Performance Of Turbo Code Over Inter-Symbol Interference Channel With Additive White Gaussian Noise

Since 1993, much interest of the coding community has focused around turbo codes, whose bit error rate is very close to Shannon limit. Since turbo codes were introduced in error correcting code field, much researches have been done to find a way to apply turbo codes to various fields of communication engineering and telecommunication systems.Present framework first reviews channel coding theories and turbo convolution codes then investigate practical performance of turbo convolutional code over noisy inter-symbol interference channels.We start in chapter 1 by considering the assumptions and conditions under which this work will be held. Reviews of inter-symbol interference channels characteristics based on the information theory concept and channel coding as well as the general treatment of soft equalization-decoding for channels that are corrupted by inter-symbol interference is given in chapter 2. Basics notions of code trellis and digital modulations are also review in this chapter. Elementary notions of turbo coding and turbo decoding are given in chapter 3. Of particular importance is that this chapter also reviews turbo-trellis coding and turbo iterative decoding. Armed with channel equalization and turbo coding-decoding notions, we propose an alternative optimal turbo equalization-decoding method in chapter 4. The proposed equalization-decoding approach works with finite inter-symbol interferences. Specifically the proposed method provides a general treatment of turbo equalization-decoding for combined source and channel coding to evaluate the performance of turbo convolutional codes over noisy inter-symbol interference channel. To come up the inter symbol interference and added white noise, higher spectral efficiency coding scheme using turbo codes and 8-PSK (or 16-QAM) is adopted.The proposed method in this framework called soft output equalization-decoding algorithm (SODA) is derived under the constraint of fixed decision delay for detection of m-ary digital signals in the presence of inter-symbol interference and additive white Gaussian noise.The Performance of the SODA has been evaluated trough computer simulation in chapter 5. Implementation and tests have been carried out using matlab. The results are obtained for channels described by the conditional probability that the constitute turbo convolutional codes are linear. Furthermore, the assumptions about the physical channel hold. That is, the channel is assumed to be stationary, band limited and having 10 inter-symbol interferences. It has been also assumed that the probability density function of the additive white noise and perfect channel estimation are available at the receiver. The simulation results are carried out for turbo code rate 1/2 with 16-QAM and rate 2/3...