Seminar: Modern Flow-based Networks for a Connected Cyber-Physical World

Ruozhou Yu

Arizona State University, School of Computing, Informatics and Decision System Engineering

February 11, 2019

11 am

Patrick F. Taylor Hall, Room 3107

Abstract

The ability to connect with everything and everyone is perhaps the most basic demand of people in this digital era. While we have built a large compilation of complex, heterogeneous and large-scale networks to fulfill this demand, it has long been observed that our current networking protocols and systems often fail to provide guaranteed performance in terms of throughput, quality-of-service, availability, load balancing, etc. This has largely impeded the practical adoption of many novel networking technologies in critical modern applications, such as the Internet-of-Things (IoT), layer-2 blockchain networks, etc.

In this talk, I will focus on my efforts in performance-oriented networking for new network technologies, using network flow-based methods. On one hand, I will show how to utilize centralized network intelligence to achieve rigorous network performance guarantee for IoT applications. Concerning different application scenarios in IoT, we propose different models along with provable flow-based algorithms to achieve various performance goals. On the other hand, I will also show how to apply flow-based algorithms in distributed and dynamic networks, by proposing a fast payment routing protocol for layer-2 payment channel networks in blockchain. This distributed protocol is able to achieve guaranteed successful payment in these capacity- and delay-sensitive networks. While the proposed solutions target concrete problems in certain networks, they also have applications and extensions beyond these specific contexts.

Bio

Ruozhou is a PhD candidate in the School of Computing, Informatics, and Decision System Engineering at Arizona State University. His research expertise lies at the intersection of networking, distributed systems, and cybersecurity. His current research interests include IoT security and networking, cloud and edge computing, blockchain-based systems and networks, big data computing and analytics, etc. His work has been published on top-tier conferences and journals such as IEEE INFOCOM, IEEE JSAC and IEEE/ACM ToN. He has served as reviewers for top journals such as IEEE JSAC, TMC, TPDS, TWC, etc.