We design and implement systems targeted at the Internet. All our projects have been driven by two major questions:
- How do we make the Internet easy to manage? This question is motivated by the fact that designing and managing large networks often involves complex policies (e.g., security, QoS, fault tolerance). Design today is largely ad-hoc. By many estimates a large portion of the IT budget of organizations is driven by the need to manage networks, with configuration errors and design faults often accounting for a large fraction of cyber-attacks.
- How do we ensure high quality of experience for Internet applications? On the one hand, latency of web applications is critical, and by many estimates even hundreds of milliseconds of latency can impact business revenue and user engagement. On the other hand, we are seeing a rapid proliferation of video traffic on the Internet with the rapid growth of Internet Television, with high quality video necessary for user retention. Yet, achieving high quality of experience is challenged by the proliferation of mobile devices and cellular technologies, while technologies such as cloud computing offer new opportunities.
Tackling these questions leads us to work in and make innovations in a number of areas such as Network Verification and Synthesis, Software Defined Networking, Cloud Computing, Content Delivery Networks, and Mobile Systems. Our projects involve designing new algorithms and systems, and often involve use of tools such as optimization and machine learning. Yet, we place a central emphasis on working prototypes, empirical methods, and data-driven analysis. Our research has benefited by support from NSF, Cisco, Google, NetApp, AT&T, and Microsoft , and many of our projects have involved collaboration with these organizations. Many of the challenges we address are motivated by real-world experience, require insights into operations of networks at scale, are great fun, and can change the world!
December 2016: Our paper, Robust validation of network designs under uncertain demands and failures has been accepted to appear at USENIX NSDI, 2017 [PDF].
April 2016: Our paper, Reducing latency through page-aware management of web objects by Content Delivery Networks has been accepted to appear at ACM SIGMETRICS, 2016 [PDF].