Iterative receivers and their graphical models

Abstract

In a number of communication systems, optimum receiver design requires joint demodulation and decoding. The complexity problem arising in practical implementation has led to an interest in iterative receivers. This chapter introduces iterative receiver algorithms based on their graphical models. These consist of representing the factorization of a function of several variables into a product of functions of a lower number of variables. Using this representation, efficient algorithms are derived for computing the a posteriori probabilities to be used for optimal symbol-by-symbol detection of the transmitted data. The unified approach presented here allows one to observe how seemingly different transmission schemes share many common features, and hence solutions that were devised for one problem can easily be adapted to a different one. Our presentation is tutorial in nature and relies heavily on graphical descriptions. © 2009 Springer-Verlag US.