Characterization and computation of optimal distributions for channel coding

Monday, June 27, 2005 - 2:00pm - 3:00pm
Sean Meyn (University of Illinois at Urbana-Champaign)
This presentation concerns the structure of optimal codes for stochastic channel models. An investigation of an associated dual convex program reveals that the optimal distribution in channel coding is typically discrete. Based on this observation we construct a new class of algorithms is introduced, based on the cutting-plane method, to generate discrete distributions that are optimal within a prescribed class. This lecture builds upon a tutorial lecture to be presented by Meyn prior to the workshop at IMA.