“DRESS Code for Cloud Storage”
Science 607-608, Byblos campus
The Electrical and Computer Engineering Department is hosting a seminar by guest speaker Dr. Salim El-Rouayheb entitled:“DRESS Code for Cloud Storage”.
The lecture will focus on how in today’s world we expect seamless access to data in the “cloud” (social networks, videos, shared documents, etc.) anywhere and anytime. Companies like Google, Amazon and Facebook, provide such services by storing petabytes of data in data centers all over the globe. These data centers use inexpensive hardware (servers, hard disks, etc.) that fails frequently. And, to quote Google, at such a large scale “failure is the norm rather than the exception”. Three-times (3x) replication has been the industry standard to achieve data reliability. However, this solution does not scale well with the exponential increase in the data. Erasure codes, such as Reed-Solomon codes, can achieve the same reliability levels with a lower storage overhead. However, they result in other system costs such as higher repair bandwidth, disk reads, latency and even security risk. But, what are the fundamental tradeoffs among the different system resources (storage, bandwidth, disk I/O, latency, energy, computational complexity, etc.) in order to meet a targeted level of data availability and security?
In this talk, Dr. El-Rouayheb will describe recent progress towards answering this question and present new codes for distributed data storage, called Distributed Replication-based Exact Simple Storage (DRESS) codes. DRESS codes permit fast data recovery upon failure with minimum bandwidth, disk reads, and computational cost at the price of a minimal storage overhead. He will describe the role of DRESS codes in achieving data security against adversaries. Moreover, he will be detailing the different projects and PhD opportunities within his research group at the Illinois Institute of Technology (IIT).
Dr. El-Rouayheb is an assistant professor in Electrical and Computer Engineering at the IIT, Chicago. He was a research scholar at Princeton University (2012-2013) and a postdoc at UC Berkeley (2010-2011). He received his Ph.D. degree in Electrical Engineering from Texas A&M University in 2009. He was awarded the Texas Telecommunication Engineering Consortium (TXTEC) Graduate Fellowship (2005) and the Charlie S. Korban award for outstanding graduate student (2004). His research is in the area of information theory and coding theory with a focus on data reliability and security in distributed storage systems.