Algebraic Methods In Distributed Storage

Refreshments will be served at 4:00 PM.

ABSTRACT

With the advent of cloud storage, big data, and social media, it is important to store user files across many computers/drives in such a way that guards against data loss/corruption. However, a greater resilience against data loss requires more storage space on the network, forcing a compromise. Furthermore, if one or many of the computers storing the file is lost, it is crucial to develop efficient algorithms to replace those missing nodes while minimizing, for example, computational complexity of the repair and/or network traffic. In this talk, I will survey some of the ways in which algebraic methods, such as algebraic coding theory, may be employed to achieve these goals, with a special focus on regenerating codes and Reed-Solomon codes. If time permits, I will report on my joint work with Hiram Lopez (Virginia Tech) on the exact repair problem for generalizations of Reed Solomon codes and possible applications to other areas of information theory and theoretical computer science.

BIOGRAPHY

Daniel Bossaller is Assistant Professor of Applied Algebra at the University of Alabama in Huntsville. He received his PhD from Ohio University under the direction of Sergio Lopez-Permouth. Professor Bossaller's research interests include applications of ring theory and linear algebra to algebraic coding theory.

A campus map is available at https://tour.stevens.edu. Additional information is available at https://web.stevens.edu/algebraic/.

Zoom Link:

https://stevens.zoom.us/j/93228680142 (Password: ACC)