EL 933 Network Measurement and Traffic Engineering
Instructor: Yong Liu
3 Credits

Room RH702


Overview

Networks, and the Internet in particular, have experienced exponential growth, becoming one of the largest engineering systems ever constructed. Management of this large system is of critical importance, and challenging at the same time. Measurements and models based on these measurements are integral components to understand network behavior and provide basis for better management. This seminar covers topics both in measurement, ranging from traffic/protocol/application characterization to anomaly detection, and in traffic engineering, including congestion control and optimal routing. The follows is the list of subjects to be covered:
  1. Internet Traffic Analysis: self-similarity and long-range dependency

  2. Packet Loss/Delay Characterization: statistical models

  3. Packet Probing: capacity available bandwidth estimation

  4. Network Tomography: inference from the edge

  5. Anomaly Detection: worm, DDoS attacks, link failures

  6. Traffic Matrix Estimation: existing and new approaches

  7. Distributed Measurement: network of network sensors

  8. Network Routing: optimal routing in a changing world

  9. Overlay Routing: application level routing adaptation

  10. Congestion Control: a closed-loop analysis


Prerequisites 

EL536 or equivalent, elementary probability course


Grading

Students will be evaluated by class participation and a final project. The final project can be either a study of a paper from major networking conferences and journals or student’s original work on network measurement and traffic engineering. For both cases, a presentation and a term paper is required.
A list of conferences:

Internet Measurement Conference (IMC)
Passive & Active Measurement Workshop (PAM)
ACM Special Interest Group on Data Communications (SIGCOMM)
IEEE Conference on Computer and Communications (INFOCOM)
ACM International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS)
IEEE International Conference on Network Protocols (ICNP) 


Reference Books

  1. “Data Networks (2nd Edition)”, D. Bertsekas and R. Gallager.

  2. “Computer Networking, a top down approach featuring the Internet (3rd edition)”, James F. Kurose and Keith W. Ross.

  3. “Nonlinear Programming”, Dimitri P. Bertsekas.

  4. “The Mathematics of Internet Congestion Control”, R. Srikant.


Course Material

Class 1:   Introduction slides in pdf

Class 2:   Internet Traffic Analysis slides in pdf
Class 3:   Network Status Characterization: packet loss and delay slides handout in pdf
Class 4:   Packet Probing: capacity, available bandwidth and bottle-necks slides handout in pdf
Class 5: Network Tomography: end-end inference of network characteristics slides handout in pdf Class 6: Anomaly Detection: worm, DDoS attacks, link failures slides handout in pdf Class 7: Selected Topics from IMC 2005
Class 8:  Traffic Matrix Estimation slides handout in pdf Class 9:  Internet Traffic Engineering slides handout in pdf Class 10:  Network Congestion Control slides handout in pdf Class 11:  Design and Analysis of Overlay Networks slides handout in pdf Class 12: Student Presentations Group I
Class 13:  Student Presentations Group II,  
Class 14:  Student Presentations Group III,  
Class 15:  Student Presentations Group IV, 


Contact

Office: LC 258
Phone : (718) 260-3959
FAX:    (718) 260-3906
yongliu@poly.edu