Naor’s model

02Nov07

We just saw our first example of a simple interaction between game theory and queueing theory. Naor’s model takes probably the most basic model from the two setting and combines them. But, I think we could already learn a lot from the comparison.

One question that came up is that of “how far the model has been extended since 69?” This is something I don’t have a good answer to and might be a good topic for a future meeting. Chapter 2 gives a number of pointers to extensions, but I’m sure there are also extensions beyond those. Is anyone interested in giving such a presentation? …going over some of the results from extensions of Naor’s model and describing some of the techniques used?

Another topic that came up was understanding the comparision between the profit max, social opt, and nash in the model. Things like the PoA…or at least a behavioral comparison.

Lastly, I noticed that some people weren’t too comfortable dealing with Markov chains. They’re fundamental to analyzing queueing models, so it’s worth refreshing yourself if you’re not confident with them. If this is you, check out the notes from my advisor’s class here:

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

Next week we’ll spend a little more time on some observable games like Naor’s model, and then start talking about unobservable games.

Filed under: Uncategorized   |  Leave a Comment