Seth Gilbert | |
Distributed Programming Laboratory
|
|
|
In general, I study algorithms for highly dynamic, distributed environments, such as mobile ad hoc networks, wireless sensor networks, and peer-to-peer systems.
These days, I am particularly interested in problems that arise in wireless networks. For example, can we communicate reliably in a collision-prone, unreliable wireless network? How about when there are malicious and disruptive devices attempting to prevent communication? One of the most significant challenges introduced by ad hoc networks is coping with the unpredictable deployments, uncertain reliability, and erratic communication exhibited by emerging wireless networks and devices. The goal of the Virtual Infrastructure Project (see also here), the topic of my PhD thesis, is to develop a set of algorithms that address these challenges and simplify the design of software and services for wireless ad hoc networks. A virtual infrastructure is a geographic abstraction that provides reliable and predictable components in an unreliable and unpredictable environment.
I am also interested in other algorithmic and complexity-related issues that arise in the study of distributed algorithms, for example, dynamic distributed storage, cache-oblivious algorithms (in particular, B-trees), and quorum deployment.