DSP applications: Signal and image processing with belief propagation

A wide range of efficient algorithms for statistical learning and interference can be applied once a problem has been formulated using a graphical model. This model provides a unifying framework for the design and analysis of error correcting codes. Effective interference algorithm also known as belief propagation (BP) propagates information throughout a graphical model via a series of messages sent between neighboring nodes. The BP was derived as an exact interference algorithm for tree-structured graphical models. It is essentially distributed variant of dynamic programming. However, BP may fail to converge in graphs with cycles, and different message update rules have been proposed allowing faster convergence.