Turbo Product Code (tpc) Repeated Soft Decoding Algorithm

Recently, much interest of the coding community has focused around a new family of error-correcting codes, called turbo codes, whose perfor-mances in terms of bit error rate (BER) are close to the Shannon limit. Since turbo codes were introduced by Berrou [1] in 1993, many papers on turbo coding have been published. The turbo codes proposed by these authors can be broadly divided into two major types: turbo product codes (TPC) and turbo convolutional codes (TCC).The goal of this paper is to describe the main ideas behind both the TPC and TCC.Although most research until now has been focused on TCCs, TPCs have been shown to be a more attractive option for a wide range of applications. TPCs are definitely more efficient than TCCs for high code rates. Another advantage of TPCs is that the performance of simple row/column interleaving is as good as random interleaving. This allows the use of simpler interleaving structures and hence a saving in terms of system complexity. In TPCs, serial concatenation is more advant-ageous over parallel concatenation and so for all practical purposes TPCs involve iterative SISO decoding of serially concatenated block codes separated by a simple row/column interleaver.In this paper, We aim at the research to the decoding Algorithm of turbo product codes, according to the principles of the Log-Likelihood Ratio and the Maximum A Posteriori Criterion, bunt the Log-Likelihood Ratio approximating expression, form the correct strategy of iterative soft decoding, and present the optimal algorithm through software. For practical implementations, simpler decoding algorithms have been suggested. This decoder implements iterative decoding of TPCs using a soft-input hard-output decoder ,followed by reliability calculations to obtain oft-decisions from the hard-output of the decoder. The performance of this decoder is nearly optimal and thus the decoding strategy presents a good compromisebetween performance and complexity, making it very attractive for practical applications. Thereafter, we present results of simulations performed during our research and a comparison of these results with the results given in .