An EXIT-Based Design Method for LDPC-Coded Schemes without Gaussian Assumptions

It is known that low-density parity-check (LDPC) codes are widely used to provide capacity-approaching performances in various transmission systems. However, the Gaussian assumption for soft messages in conventional EXIT (extrinsic information transfer)-chart-based optimization algorithms result in prediction gaps in the decoding threshold for certain transmission systems. In this letter, we propose an EXIT-chart-based optimization method, in which all the Gaussian assumptions for soft messages can be avoided. Compared to the conventional design methods, the proposed method provides a more accurate threshold prediction and can be used to design LDPC-coded schemes with better error performances.