Title: "Channel Capacity: From Waves to Particles and Back Again"
Abstract: The well-known Shannon capacity expression for the bandlimited additive gaussian-noise channel is correct from the point-of-view of the mathematics, but incomplete from the point-of-view of the physics. This is because of the emergent granularity of a lightwave at low signal levels. Quantum information theory, though formally the proper tool to study this, is a tool too sharp to obtain insightful answers. We propose an intermediate semiclassical information theory based on the Poisson trans-form of Mandel and Wolf. The wave and particle views of a lightwave are seen to be the two sides of the Poisson transform. The capacity of a photon channel conjectured by Gordon (1962) and Forney
(1963) based on maximum-entropy considerations is proved to be correct, and the Shannon capacity of the bandlimited gaussian noise channel is shown to be the emergent capacity formula as the number of photons becomes large.
Biography: Richard E. Blahut was Professor of ECE at the University of Illinois from 1994 to 2014 and Head of that Department from 2001 to 2008. He is now at the University of Penn-sylvania as an Adjunct Professor. Professor Blahut’s own research pertains to coding theory and algorithms for signal processing and image formation. While at IBM, he pioneered pas-sive coherent location systems, which are used for U.S. Department of Defense surveillance systems. He also established error-control codes that have been used in the high-speed telecommunications systems for military helicopters and long-range cruise missiles. That research resulted in his first textbook, Theory and Practice of Error Control Codes
(Addison-Wesley), which was published in 1983, while he was simultaneously teaching as a courtesy professor at nearby Cornell University, his doctoral alma mater. Over the ensuring decades, he has continued to publish prolifically. His 11th book is entitled Cryptography and Secure Communication.
