Comments on our first meeting

23Oct07

Sorry I couldn’t make it to our first meeting, but I hear that it went well. If you have any questions or comments on the background Mani went over, just respond to this post.

Also, a question for you guys:

Mani didn’t get to do much with Markov chains. How comfortable is everyone with solving simple chains?

As a test, suppose we have an M/M/1 queue (exponential interarrival times and job sizes). If we make the state of the chain the number of jobs in the system, does everyone know how to setup and solve this chain for the stationary distribution of number in system?


Filed under: Uncategorized   |  1 Comment

One Response to “Comments on our first meeting”  

Feed for this Entry Trackback Address
  1. 1 Adam on October 23, 2007 said:

    For those who aren’t too familiar with queueing and markov chains, there are some excellent notes available here:

    http://www.cs.cmu.edu/~harchol/Perfclass/NotesFall07/notes.html

    -The Poisson process lecture presents three equivalent definitions of the process that may help you to gain intuition.
    -The M/M/1 lecture will walk you through the answer to the question I asked in the post
    -The Little’s Law lecture gives some nice applications of the result
    -I don’t think Mani covered PASTA (Poisson Arrivals See Time Averages), and this is another fundamental queueing result.


Leave a Reply