Achievable Information Rates of Bit-Interleaved Coded Modulation

Although widely used in the design of iterative systems, EXIT charts lack of operational meaning in terms of achievable rates. This project aims at providing such an operational meaning by using a mismatched decoding framework. Mismatched decoding theory allows for the natural modelling of iterative decoders, by which the information generated at each iteration modifies the mismatched decoding metric, and accordingly the achievable rates. More specifically, this project is aimed at exploring the precise way in which the achievable rates so determined can be related to common tools in iterative decoding analysis, such as density evolution or EXIT charts. We will apply the results to multiple-antenna, additive exponential noise, additive energy and discrete-time Poisson channels.