COURSE DESCRIPTION
Have you ever wondered how services like Netflix stream video so smoothly, why your online game is lagging, or how the global network of routers and cables that make up the Internet is actually structured? This course moves beyond the textbook theory of computer networks and dives into the practical, hands-on science of Internet measurement . We will explore the tools and techniques used by network engineers, researchers, and major tech companies to monitor, diagnose, and understand the health, performance, and evolution of the global Internet. This is a lab-focused course where you'll get your hands dirty every week. You won't just learn about fundamental tools?you will master them. Through a series of carefully designed labs, you will capture and dissect packets, launch global-scale measurements, analyze speed test data, and map the Internet's topology.Upon successful completion of this course, you will be able to:
- Analyze Network Traffic: Use Wireshark to perform deep packet inspection and diagnose network protocol issues at a professional level.
- Deploy Active Measurements: Design and run ping and traceroute measurements from thousands of global vantage points using the RIPE Atlas platform.
- Interpret Performance Data: Understand the mechanics behind speed tests and analyze open data from the Measurement Lab (M-Lab) consortium.
- Map the Internet: Utilize large-scale datasets like the Internet Yellow Pages to explore the Autonomous System (AS-level) topology of the Internet.
Course Structure
The course is structured around weekly modules. Each week has a 3-hour Monday session (a short mini-test on the week's reading, a debrief, and a 2-hour hands-on lab) and a 1-hour Wednesday session where the teaching assistants walk through that week's lab.
- Weekly Labs: The core of the course. You will work directly with powerful, real-world measurement platforms and datasets every week.
- Guest Lecture from Industry Experts: We are thrilled to host an exclusive guest lecture from senior engineers and researchers at a leading measurement organization such as RIPE NCC or M-Lab. This is a unique opportunity to learn directly from the people who build and maintain the measurement infrastructures we study.
Prerequisites:
A solid understanding of computer networking fundamentals (e.g., equivalent to CS-335) is required.
Evaluation
There are no midterms and no final exam. Your grade is built up week by week, allowing you to focus on mastering the concepts and skills as we cover them:
- Theory (40%): Short mini-tests at the start of the Monday session, on the reading assigned beforehand. There are ten; the first is a practice run that does not count, and your best seven of the other nine count.
- In-lab work (40%): The work you submit during each Monday lab. Your lowest two marks are dropped.
- Take-home exercises (20%): One per lab, due Sunday evening. They are assessed through questions in the next day's mini-test, and you earn those marks only if you submitted the exercise.
- Bonus project (up to +5%): An optional blog post, added on top of your grade. Details will be announced later in the term.
There are no makeup mini-tests; the dropped marks are there to cover missed weeks.