Logistics
===============
* Website: https://cs.nyu.edu/~anirudh/CSCI-GA.2620-001/
* Instructor: Anirudh Sivaraman
* Timing: Thursday 7:10 pm to 9 pm
* Online: https://nyu.zoom.us/j/98867434693
* Office hours: Friday 1 to 2 pm (room 408, 60 FA) or Zoom.

Class grading
===============
* Research projects in teams of up to 2 (60%): Decide on project title soon.
* Midterm exam (20%)
* Final exam (20%)
* (Will decide with sufficient notice if it's going to be in-class or take-home.)

Prerequisites
==============
* Undergrad understanding of networking, OS, and data structures
* Math: Some familiarity with probability, linear algebra, boolean logic, and differential equations

Participation is important
==============
* It's a grad class and there aren't well-defined solutions (or problems) that I can just repeat to you.
* It makes the class more fun---especially if you disagree with the papers.
* I'll get a sore throat if I lecture for two hours straight.

Lectures
=============
* 1 hour and 50 minutes
* Will break up into two halves with a 5 minute break in between: 7:10 to 8 or 8:05 and 8:05/8:10 to 9
* Remind me to take a break in between if I don't.
* Mostly Q&A style. Read the papers in depth and come prepared with questions, concerns, critiques, etc.
* Mostly whiteboard; might use the projector when appropriate

Paper reading
=============
* Should take 2 to 3 hours per paper. Do two passes over the paper to get the material.
* Will highlight sections of the paper that can be ignored (if such sections exist).
* Read Keshav's "How to Read?"
* In general, be critical, but not cynical.
* Be positive and try and find out what is interesting about the paper ...
* but also ask critical questions; research is never complete.
* Thinking critically:
-> Are they making the right assumptions?
-> Whose problem does the paper solve?
-> Is the solution complete or is it a first step?
* Paper selection is informed by my taste, but is intended to cover a broad range of networking topics

Projects:
=============
* Start thinking about project topics right away.
* Major portion of your class grade.
* Any topic that intersects with networking (e.g., building a new cryptocurrency on the Internet)
* But, there must be a significant research component (i.e., "An implementation of X" is not sufficient)
* Think big for your project, but act small in terms of week-to-week tasks for your project.
* And meet me regularly for feedback, narrowing down scope, etc.
* Decide on project title two weeks from now
-> First, look at the lecture topics.
-> Then, think about what interests you and how to extend the lecture topics into research projects.
-> Research at the intersection of networking and other areas is especially encouraged.
-> Then, come talk to me.
-> Please start early! It always take longer than you expect.
* I expect many of your projects to result in papers with additional work, so aim high.
* Project checkpoint: Present preliminary results.
* Final poster presentation.
* 6-page report in two column format due with final presentation

Reproducible Research Projects:
==============
* Projects can also be around reproducing and siginicantly extending prior published papers.
* But will expect the same amount of effort regardless of which of the two options you pick.

Forum:
==============
* Campuswire: You should have all received an invite.
* Will use it for announcements.

Final and midterm exam:
=============
* Details to be announced later
* Everything on the papers and everything discussed in lecture is fair game on the exams.

Notes:
============
* You can use the bullet points and previous year scribe notes.
* Also, please ask on the course forum if anything's confusing.