Optimizing the parameters of turbo product codes using genetic algorithms

When selecting an error correcting code, it is desired to fulfill a data error rate criterion, but also the code that is selected does this without being excessively complicated. For specific channel conditions it is quite difficult to optimize the error correcting code parameters' analytically. This work proposes multi-objective optimization by applying the Genetic Algorithm (GA) in the selection of Turbo Product Codes (TPC) parameters' that are used for transmission of data over an AWGN channel. The results show that the GA is capable of converging to a set of sensible solutions and giving the pareto-optimum set for error performance against code complexity.