Reducing the convergence loss of 3-dimensional turbo codes

This paper deals with the convergence loss reduction of a 3-dimensional turbo code combining both parallel and serial concatenation. This code is derived from the classical turbo code by concatenating a rate-1 post-encoder at its output. In order to reduce the observable loss of convergence at high error rates, we propose first a time varying construction of the post-encoded parity. Then, we analyse the association of the 3-dimensional turbo code with high order modulations, where both the systematic bits and the post-encoded parity bits are more protected than the other parity bits. Performance comparisons are made between the 3GPP2 standardized turbo code and the 3-dimensional 3GPP2 turbo code. The different stages are illustrated with simulation results, asymptotical bounds and EXIT charts.