Non-Binary LDPC Code Optimization for Modern Storage Systems




Abstract: Transmission channels underlying modern dense storage systems significantly differ from canonical channels. While existing LDPC codes optimized for symmetric channels are being actively considered for Flash applications, we demonstrate that such approaches are inadequate. We propose the weight consistency matrix (WCM) framework to analyze and optimize non-binary LDPC codes for non-canonical channels. We show significant performance gains on asymmetric Flash channels using the WCM framework.

Bio: Ahmed Hareedy is a Ph.D. candidate in the Electrical Engineering Department at UCLA. He received his B.S. and M.S. degrees from Cairo University in 2006 and 2011. His research interests include information theory and coding theory. He worked with Mentor Graphics Corporation between 2006 and 2014. He worked as an ECC architect with Intel Corporation in 2015. He is a recipient of the best paper award from GLOBECOM 2015. He will give a graduation day talk representing UCLA at ITA 2017.