This research is aimed at addressing several key issues in the theory and designs of rateless codes for wireless communication applications. Unlike conventional coding designs, a rateless code does not require a prescribed transmission data rate. Instead, the encoded messages are transmitted until the receiver signals that it has successively decoded the message. This technique allows the data transmission to adapt to the realized channel quality, which is often not known in wireless network applications. The main approach used in this research to achieve rateless transmissions involves the use of multiple layers of low rate sub-codeswith successive cancellation receivers. Related issues in the fundamental performance limits and specific code designs are studied.
Specifically, the investigators examine the rateless code designs in additive white Gaussian noise (AWGN) channels with unknown signal-to-noise ratios, by using repetition codes and dithering encoding techniques. The approach is then generalized to study binary symmetric channels with unknown capacities, multiple-input multiple-output (MIMO) channels, as well as network applications including multiple access channels, relay channels, as well as a variety of cooperative transmission architectures in the rateless context. The ultimate goal of this research is to addressthe fundamental issue of incremental information transmissions, and to develop a theoretical foundation for the development of robust and flexible wireless infrastructure needed for a new generation of applications.
