Thank you for thinking of NAL!
PhD semester projects
Contact: Katerina.
MS theses (PDMs) and MS/BA semester projects
Our long-term goal is to improve Internet transparency, and we keep looking for new ways to achieve that. At the moment, we have projects on the following topics:
- Using online games to map Internet latency: We collect video-game streaming footage from Twitch and extract the latency numbers to build a universal map of Internet latency. We are looking for a motivated student to help us analyse the collected measurements: find gamers with interesting latency patterns; group gamers according to their latency behavior; and try to explain natural variations in latency due to real-world events [MS or advanced BA project]. We are also looking for a motivated student with good experience in Natural Language Processing (NLP) to help us analyze gamers’ reactions to latency changes [MS or advanced BA project]. Contact: Catalina
- Using math to capture non-neutral network behavior: Network devices may behave “non-neutrally,” i.e., treat some traffic flows worse than others. This may happen by accident (because of a bug or misconfiguration) or on purpose (because of an explicit policy). We want to define a mathematical metric that formally captures such non-neutral network behavior. We think such a metric would compare: (a) the per-flow congestion observed at communication endpoints, e.g., through Explicit Congestion Notification (ECN) signals; and (b) the overall congestion measured at network devices. If a network behaves “neutrally” (treats all flows the same), these two quantities should “align.” We are looking for a motivated student to help us develop (a) a simulator of different data-center topologies and configurations, as well as (b) an ECN-data analyzer that can distinguish between non-neutral network behavior and normal network-performance fluctuations [BA project]. Contact: Mahdi
- Using edge caching to map Internet bandwidth: Most content that end-users download from the Internet today comes from edge caches (that are managed by public or private clouds). By measuring the cache hit rate achieved when downloading different content from different caches, we can infer the approximate popularity of each content in each area of the world. This, in turn, can give us information on the bandwidth available to different end-users (e.g., end-users with more bandwidth tend to download higher-quality videos). We are looking for a motivated student to help us measure cache hit rates from different vantage points around the word and use the measurements to create a map of Internet bandwidth [BA project]. Contact: Abdullah
To enjoy any of these topics, you must have enjoyed your Computer Systems and/or Networks courses (including all those quirky delay computations and TCP diagrams).
Before we start working together, it’s important for both you and us to agree that the project matches your interests and background. So, if you are interested, please contact us to schedule a meeting.
Cybersecurity MS
The above topics do not qualify as security or privacy topics, so they are not good matches for Cybersecurity MS semester projects. Occasionally we do offer such projects, but unfortunately this won’t be the case in Fall 2024 and Spring 2025.