Turbo coding is the conventional name given to an exciting and relatively new digital communication technique capable of approaching the theoretical limits imposed by Shannon theory. The technique has enjoyed enormous recent attention from coding theorists, but many practical, engineering-oriented issues remain before these coding techniques will see widescale practical implementation. We propose research in the areas of: (1) reduction of decoder latency and complexity; (2) developing stream encoder and decoder architectures; (3) efficient synchronization at low SNR; (4) improving performance of small-frame turbo systems; (5) robustness to channel modeling error, either in estimation of SNR or non-Gaussian noise; and (6) use of turbo codes on non-coherent or differentially-coherent channels.
