Cellular Interference Management and Spectral Graph Learning
|Event Date:||April 9, 2015|
|Speaker:||Professor Aly El Gamal|
|Speaker Affiliation:||University of Southern California, Ming Hsieh Department of Electrical Engineering
|Contact Name:||Professor David Love
|Open To:||ACCEPTABLE FOR ECE 694
Abstract: The first part of this talk is on novel designs for the cellular backhaul that facilitates interference management. Enhancements in the cellular infrastructure are anticipated for future generation cellular networks. We explore the potential benefit of having a flexible backhaul design where network connectivity and statistics of deep fading conditions are taken into account. For the cellular downlink, we study assignments of messages to base station transmitters and cooperative transmission schemes that deliver the promise of interference alignment while meeting delay requirements. In particular, we highlight the importance of basing the decisions for cell associations as well as assignments of messages over the backhaul on the knowledge of the structure of dominant interfering links. At the end of this first part, we discuss dual designs in a cellular uplink model where the base station receivers can share decoded messages through the backhaul. In the second part of the talk, I will present recent results on analyzing the performance of spectral graph methods for semi-supervised learning. The problem of semi-supervised learning is gaining rising interest in the context of big data analytics where the amount of unlabeled data far exceeds available label information. This problem has applications in learning network channel state information when the coherence time is small. Graph-based methods have been quite successful in solving the semi-supervised learning problem, as they take into account the underlying geometry of the data. For distance-based similarity graphs, we analyze the asymptotic performance of the recently introduced method of band-limited interpolation of graph signals. Our result strengthens the notion that learning using distance-based similarity graphs is akin to solving the low density separation problem on a finite sample.
Biography: Aly El Gamal received the B.S. degree in Computer Engineering from Cairo University, Cairo, in 2007, the M.S. degree in Electrical Engineering from Nile University, Cairo, in 2009, the M.S. degree in Mathematics and the Ph.D. degree in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign in 2013 and 2014, respectively. He worked as an intern at the Office of the Chief Scientist of Qualcomm Inc. in 2012. He joined the Ming Hsieh Department of Electrical Engineering of the University of Southern California as a Post-doctoral Research Associate in 2014. His research interests include information theory and wireless communications, learning theory and big data analytics, and graph theory.