Best Paper for Work on Regenerating Codes
A paper co-authored by University of Illinois computer science professor Brighten Godfrey received the 2010 IEEE Communications Society Technical Committee on Data Storage Best Paper Award.
The award is given by the Signal Processing for Storage Technical Committee of the IEEE Communications Society to a conference or journal paper published in 2010 that presents novel research on data storage.
The paper, "Network Coding for Distributed Storage Systems", was authored by Alex Dimakis, Brighten Godfrey, Yunnan Wu, Martin Wainwright, and Kannan Ramchandran, and appeared in the IEEE Transactions on Information Theory.
The paper created a new sub-area of coding theory called regenerating codes, which can make reliable storage in data centers more efficient.
Large data stores, which often appear in today's data centers and cloud services, use erasure coding to provide redundancy across multiple disks so if part of the file is lost it can be reconstructed. But the erasure-coded fragments must also be replenished after a failure, to maintain the same level of reliability. Traditionally a lost fragment is replaced by first reconstructing the entire file, and then distilling it into a small additional fragment to replace the lost fragment. The paper introduced a technique called regenerating codes which can produce new fragments much efficiently, without reconstructing the entire file.
The award was presented at the IEEE Global Communications Conference (Globecom) 2011 in Houston, Texas, in December.
Alexandros G. Dimakis, P. Brighten Godfrey, Yunnan Wu, Martin J. Wainwright, and Kannan Ramchandran. Network Coding for Distributed Storage Systems. IEEE Transactions on Information Theory, vol. 56., no. 9, September 2010