CS 536 Term Project (Spring 2006)

Project theme
------------- 
Tool based and quantitative analysis of the Internet

Introduction
------------
There is tremendous interest in understanding the existing Internet given
its success and its interesting properties as a complex and large scale 
system. The Internet is live and operational, supporting real applications
by real users. This gives excellent potential for understanding the Internet
by empirical studies.

Students can work in teams of up to two persons. The projects are somewhat 
open ended, but general approach (progressive from more basic to more 
advanced) to completing your project is the following:

1) Read a few papers on a selected topic. The paper collection should
include those that are empirical in nature, including modeling based on
empirical data. (Further paper references on selected topics will be 
available from the instructor.)

2) Read up on the corresponding data repositories, data analysis tools,
and other software supporting the empirical studies.

3) Where appropriate, obtain the data traces, data analysis tools, software 
used by the papers studied. Experiment with the data and tools to try to 
reproduce the conclusions you read in the papers. If possible, generate new 
data / observations / empirical studies using the data, tools, or software
available to you.

4) Because of the open-ended nature and customizability of the project 
topics, discussions with the instructor, including checkpointing and 
expectations of progress, are strongly encouraged. 

Possible topics 
---------------

1) Write a report on the route traces in the Oregon Routeviews 
(www.routeviews.org) and RIPE (www.ripe.net) projects. Give a survey
of the data analysis tools available and the existing research using
the data.

2) The following projects make use of the CAIDA (www.caida.org) repository
for Internet data analysis. Give a summmary of the main CAIDA data and
analysis tools available. Then research into and report on one of the
following: 

2.1) Network telescope (www.caida.org/analysis/security/telescope)
2.2) The topology and connectivity of the Internet. Possible subtopics are:
connectivity models (e.g., powerlaw), denail-of-service backscatter, and the
geographical distribution of the IP address space.
2.3) Analysis of Internet AS structures and BGP routing tables. Understanding 
of BGP instability / misconfiguration issues.

3) The following topics are in network security

3.1) Survey the existing literature on worm identification (robust signature
extraction) or modeling the spread of worms in the Internet.
3.2) Give a survey on privacy routing in the Internet. In particular,
experiment with Tor (http://tor.eff.org) and evaluate its runtime / management
overhead and its vulnerability to attacks (e.g., DDoS).

4) Wireless mesh network. Read and report on existing empirical studies
on multihop wireless networks (e.g., the MIT Roofnet project), including
the modeling of wireless links, effective routing metrics, and high throughput
routing techniques.

5) Overlay network. Install and experiment with Xen 
(www.cl.cam.ac.uk/Research/SRG/netos/xen) as a system overlay / virtualization
tool supporting standard user applications.