Decision feedback equalizers: Concepts towards design guidelines
Brian D. O. Anderson, Rodney A. Kennedy, PR Bitmead · 1st IASTED International Symposium on Signal Processing and its Applications · 1987
A quantitative and qualitative analysis of the stochastic dynamics of the blind adaptation of decision feedback equalizers (DFEs) operating on noiseless communication channels is presented. The overall goal, not yet acheived, is to identify channel classes for which DFE use is a practical option. Averaging analysis is invoked to predict the existence and location of undesirable attraction points to the adaptation algorithm. The parameter space in which adaptation takes place is shown to be subdivided into a large but finite set of convex polytopes when the channel can be modelled as a finite impulse response filter. With each polytope we show that the parameters describing the averaged equation are fixed. Further, for every point in a given polytope the error performance is described by a single finite state Markov process. Hence, the adaptive dynamics and the error properties of a noiseless DFEs do not depend continuously with the operating point in parameter space but rather only as a function of polytope membership. The possibility arises of the algorithm becoming hung at an inappropiate attraction point.