From Theory to Application: Block-Structured Integer Programming Meets Blockchain
When: Monday, February 11, 2019
Where: PGH 232
Time: 11:00 AM
Speaker: Dr. Lin Chen, University of Houston
Host: Dr. Panruo Wu
This talk will cover my recent research in the algorithmic foundation of distributed systems, and its application in the security and scalability analysis, with a particular focus on blockchain. Security and scalability are major concerns in the design of a distributed system, yielding a series of optimization problems in scheduling, routing, resource allocation, etc. Block-structured integer programming is a strong mathematical tool for modeling a broad class of such optimization problems. Recently, we have made significant progress by establishing a polynomial time algorithm for a broad subclass of block-structured integer programming. I will briefly talk about the basic idea of this algorithm. Then I will show how to utilize the theoretical results on block-structured integer programming to develop a scalable blockchain-based smart contract system, to analyze the security of a multiagent system under uncertainty, and to establish the security threshold of newly developed consensus protocols, e.g., the Proof-of-Elapsed-Time scheme proposed by Intel.
Lin Chen is a research assistant professor at University of Houston. Before joining UH, he was a postdoc at Technical University of Berlin, Technical University of Munich and Hungarian Academy of Science. He obtained his Ph.D at Zhejiang University. His research interests include algorithms and complexity, security and scalability of blockchain systems, distributed computing. His research has been published in journals such as SIAM Journal on Computing, ACM Transactions on Algorithms, Journal of Computer and System Sciences, and in conferences such as SODA, SPAA, AAAI, AAMAS. He is currently the PI of an NSF grant on approximation algorithms and applications.