Objectives

After the course, you…

  1. are able to measure and analyse basic properties of network traffic and draw conclusions on the results
  2. are able to apply statistical methods in processing, analysing, and presenting the measurement data; also able to critically evaluate the applicability of the methods
  3. understand the technical and legal issues related to network measurements
  4. are familiar with methods and tools related to network traffic measurement and analysis

Prerequisites

It’s easier to pass the course if you already know:

  • Basic knowledge of IP networks. ELEC-C7240 (or equavalent) recommended.
  • First course in probability and statistics (MS-A050x)
  • Linux command line basics
    • awk, sed, tr, grep, cut, bash
  • (Script) programming to make measurements and pre-process data
    • python, perl, javascript, java, ruby, php, C++, C#, go, bash, …
  • Statistical software like python (with numpy, pandas and mathplotlib) or R for analysis

Course personnel can best support the Linux-python toolchain, but you are free to choose the tools you like best. Try Linux on VirtualBox on Windows or OS X if you are unfamilar with it.

How to pass the course?

  • Master “Internet Traffic Measurements and Analysis” topics
  • Final assignment max 70 points – you need a passing grade (minimum points)
  • Five exercise assignments are mandatory and will give
    • Max 30 points
    • Acquire much of skills needed for final assignment
    • If you are not able to make to a some exercise event for some reason, a small extra work is required (actual assignment needs to be returned within time; extra work by December 6th)
  • Lectures on Wednesday mornings
  • Mandatory exercise/help events on Thursdays (two-hour slots, not every week) and Mondays (two-hour session, also not every week)

(Almost) Weekly exercises (5 instances)

  • Introduction on Thursday: initial group discussion and review
  • Two (or three) sessions per day: groups opened after lecture
  • Dead-line on Wednesday before next session 22:00
    • Late return: max 15 points
  • Return via MyCourses
    • If you find an error in your submission after dead-line, do NOT resubmit the fixed version before receiving acknowledgment from course staff. If you do, your submission is seen as late.
  • Review on Thursday with discussion and comments
  • These are mandatory, with option to replace no-show with additional report of an assigned subject (1-2 pages)

Access to weekly exercises

  • Will be carried out as Zoom sessions
  • Course staff will give introduction and available for helping you out
  • Remote access to classroom computers

Options for running experiments

  • Your own computer
    • Linux recommended
    • Windows users: run virtual Linux, WSL might work
    • MacOS and *BSD operating systems: beware of different command line usage
  • Aalto Virtual desktop https://vdi.aalto.fi
    • No heavy computation on virtual hosts
    • Provides full desktop via browser or VMWare Horizon Application
  • Aalto Linux servers: kosh.aalto.fi and lyta.aalto.fi for lightweight processes, brute.aalto.fi and force.aalto.fi for heavy computation
  • Aalto Linux classroom computers

Final Assignment

  • Two parts
    1. ready dataset given to analyse
    2. collect your own dataset and analyse it
  • Analyse and make a clear report. All work must be individual!
  • Dead-line by end of November sharp (2021-11-30T23:59 Finnish time)
    • Late submission gives grade 1 at best; Return MVR early, do not resubmit fix after DL (unless agreed with staff)1
  • Review discussion on Monday 2021-11-22 – you should know how to complete the assignment at this state
    • Mandatory event: if you cannot make there for some reason, contact course staff well before dead-line.

Where to get help to pass the course?

  • Excersise sessions on Thursdays 8-16, Final Assigment on 2. period on Mondays.
  • Discussions on Zulip, registration link will be published at For Aalto Users page at MyCourses
  • Peer support is encouraged but submissions must be individual
    • Plagarism is very obvious when multiple people report the same graphs although data has been different.

Material

  • Lecture notes by Markus Peuhkuri
  • Slides and extra material provided by lecturers
  • Books: (can be found from Aalto library, some as ebook)
    • Data Analysis:
      • David S. Moore and George P. McCabe, Introduction to the Practice of Statistics, 5th Edition, W.H. Freeman & Co., 2006 -> Chapters 1,2
    • Sampling and experimental design:
      • David S. Moore and George P. McCabe -> Chapters 3,5
    • Probability models and measurements:
      • Sheldon M. Ross, Introduction to Probability and Statistics for Engineers and Scientists, 5th Edition, Elsevier, 2014
      • Mark Crovella and Balachander Krishnamurthy, Internet Measurement: Infrastructure, Traffic, and Applications, John Wiley & Sons, 2006
    • Stochastic processes in network measurements:
      • Mark Crovella and Balachander Krishnamurthy (above)

Personnel

  • Lecturers
  • Assistants
    • Weixuan Jiang
    • Suvro Jyoti Kundu
  • Best way is to reach via course Zulip

  1. MVR=Minimum Viable Report↩

Last modified: Saturday, 11 December 2021, 5:39 AM