Advanced Networks Colloquium: Alex Dytso, "Treating interference as noise in competitive scenarios"
Friday, July 18, 2014
2328 A.V. Williams Bldg.
The Advanced Networks Colloquium
On the optimality of treating interference as noise in competitive scenarios
University of Illinois at Chicago
We study performance of Treating Interference as Noise (TIN) for the symmetric two user real-valued additive white Gaussian noise Interference Channel (G-IC). Surprisingly and contrary to the common believe, TIN is shown to be optimal in a Generalized Degrees-of-Freedom (gDoF) sense and to be to within O(log(log(SNR))) of known outer bounds for all channel parameters outside of an outage set of controllable measure. The key novelty is the development and analytically characterization of the rates achievable by a new strategy that uses a superposition of Gaussian and discrete random variables as channel inputs, referred to as mixed inputs.
Alex Dytso is currently a Ph.D. student in Department of Electrical and Computer Engineering at University of Illinois at Chicago (UIC) where he is working with Dr. Daniela Tuninetti and Dr. Natasha Devroye. Alex received his B.S. in Electrical Engineering from UIC in 2011. Alex is a recipient of several awards such as Wexler Award for promising graduate students and International Engineering Consortium's William L. Everitt Student Award of Excellence for outstanding seniors who have demonstrated an interest in the communications field. His research interests lie in the area of information theory, coding theory and wireless communications. Currently he is pursuing an internship position with NASA Goddard Space Flight Center.