Network Architectures

Lectures happpen on thursdays and tuesdays in FS07 at 2pm.

This is a paper reading seminar style course - there will be a schedule giving the list of work to read and who is assigned to which paper each week.

There's a great book for background reading to kickoff, by Dave Clark from MIT on network architecture(s) which I highly recomend. A fairly recent Keynote by Henning Schulzrinne gives a great persepctive on the past and recent history of networking. Even more recent is this 40+ years retrospective of the Internet

I also recommend this great talk by Scott Shenker on the future for software defined networks - essentially why computer science more than ever has relevance in communications systems.

schedule of when you will be speaking, and link to papers you'll be talking about. This schedule is subject to alteration!

Student's speaking assignments are to give a 20-25 minute prepared talk on the paper. I am very happy to discuss paper assignments and talks beforehand. Advice on critical reading/reviewing, writing, presenting a paper is offered in the Research Skills Programme

We'll go through these topics at roughly one per week.

One thing I'd like readers to bear in mind is that one can take an evolutionary approach to network architecture change, or one can try to be revolutionary. In discussing a given paper, try to see which approach it is taking and whether this supports or undermines the viability of the proposed idea - this notion originated with Constantine Dovrolis and Jenifer Rexford in this nice counterpoint discussion. An important evolutionary refinement is Punctuated Equilibrium: which may be how technology (including networks) evolve really.

A very interesting complex systems/systems bio/eco/evolutionary view on how layered architectures evolve is this paper on Architecture, constraints, and behavior by John C. Doyle and Marie Cseteb.

Interesting reading are the papers in this workshop on the impact of the pandemic on teaching networking

Forwarding/Addressing & IPv6 & The Internet Architecture for Oct 5 & 10

Course Introduction and Lecture 1

The Internet Vanilla Architecture 2

The Design Philosophy of the DARPA Internet Protocols

For IPv6 evolution, see also:

The Simple Internet Protocol (SIP) (not to be confused with the internet telephony SIP:-))
Paul's Internet Protocol
recent workshop report on future internet

Radical Alternatives to The Internet Architecture Oct 12 & 17

All students to prepare a 7 minute talk with 7 bullet points about the two papers, on Haggle and Plutarch here- well worth reading the Dovrolis paper and paying attention to Evolution v. Revolution as well. See also on concise presentations

My responses

Background for Essay 1 for Nov 6

See Essay 1

Topology/Data Centers - Oct 19 Lecture & Talk

Then One of you will cover Data Center net architectures. See azure h/w acceleration and pingmesh), versus Facebook and Google

ML and Networks Oct 24/Oct 26

Then one of you will discuss evolutionary programming for TCP, or re-enforcement learning for cellular, or NLP for disambiguating" spec and code generation

Experiments on the Automatic Evolution of Protocols using Genetic Programming is even more radical, actualyevolving the whole protocol, not just performance!

More background on the amazing range of TCP congestion control algorithms in the Internet today

I will cover Qjump and associated artefacts!

2022 cornell sigcomm paper on Tbps to the host just to show where we are now with a "single" machine!

Wireless and Mobile- Oct 31/Nov2 Lecture & Talk

One of you will tackle this paper on fingerprinting Mobile Virtual Operators: Mobile Operator fingerprinting A repository of Mobile Network Measurements for interest/background. Note this year's cloudflare connection tampering paper

I'll cover Wireless Capacity arguments

physical considerations for networks also happens in fixed networks, as in this paper on critical regions Impact of Human Mobility on Opportunistic Forwarding Algorithms describes the arbitrary delays you get with the store-carry-forward paradigm!

See also David Tse's Qualcomm lectures on wireless capacity I and II