Exploiting The Turbo Principle

Ohio State University Research Foundation -Do Not Use, Columbus OH

Investigators

Abstract

In recent years, Turbo coding has emerged as a powerful forward error control coding paradigm. Although still new, the success of Turbo coding is witnessed by its widespread use in a variety of applications, including deep space communication, satellite communication, and digital cellular communication. The key behind this success is the elegant iterative decoder that approaches theoretical limits with reasonable implementation complexity. More recently, the iterative decoder structure was abstracted and generalized to what is now known as the Turbo Principle. Because of its remarkable performance, it is expected that the Turbo Principle will have an enormous impact on virtually all aspects of digital communication systems. The main goal of this research is to explore the fundamental limits of the Turbo Principle while offering practical implementations. The first research thrust aims at developing an analytical approach for computing the convergence thresholds of Turbo decoders. In addition to the expected savings in computational complexity, the analytical approach is sought to offer more useful insights for code design. The convergence analysis will facilitate constructing irregular ``Turbo-like" codes that approach the capacity limit of Additive White Gaussian Noise (AWGN) channels. The code performance with small and moderate block lengths will be considered as an additional design criterion. This research will also investigate graphical code design for multi-path fading channels. Code constructions will be optimized for parametric fading models that capture the different limitations in practical systems. Furthermore, The convergence analysis will be used to explore fundamental limits on the performance of Turbo multi-user receivers in generalized fading channels.