Homework #9 - Scanning - Due Sunday Sep 25, 11PM


Turn in this assignment via email (vern@berkeley.edu) by the due date, with the term Homework in the Subject.

Read the paper Fast Portscan Detection Using Sequential Hypothesis Testing, Jaeyeon Jung, Vern Paxson, Arthur Berger, and Hari Balakrishnan, IEEE S&P 2004
(Note: as is generally the case, it's fine to skip the Appendix.)

Briefly write up your views of:

  1. What are the main contributions of this paper?

  2. What parts of the paper do you find unclear? (optional)

  3. What parts of the paper are questionable? (That is, you think a conclusion may be wrong, an approach or evaluation technically flawed, or data ill-presented.)

  4. Propose a detection problem other than scanning for which Sequential Hypothesis Testing would provide a natural framework for the task. (Devise this on your own; not drawing upon external resources.) Be explicit about the definition of the Y_i indicator variables, and discuss to what degree they are well-modeled as IID. Sketch the steps you would need to take to parameterize the detector - including priors - and evaluate its efficacy. Explain how well you think it would work for the particular problem, and why.