Recap: These are my opinionated reviews of papers from a course I’m auditing.
“Jeff Shneidman and David Parkes, Rationality and Self-Interest in Peer to Peer Networks, The 2nd International Workshop on Peer-to-Peer Systems (IPTPS’03)”
This paper has three goals. The first is to convince the reader that rationality is a real issue in peer to peer networks. The second is to introduce mechanism design as a tool that can be used when designing networks with rational nodes. The third is to describe three open problems that are relevant in the peer to peer setting but are unsolved in existing AMD/DAMD work. In particular, we consider problems that arise when a networking infrastructure contains rational agents.
Ok, so the first goal is irrelevant because I get it. I’m interested in the second goal. Depending on how that goes, I may or may not be concerned about the third goal.
Although it has not been labeled as such, rational behavior has occurred in computational peer to peer settings as well. One perverse example occurred when users of Seti@Home (a peer to peer computation project) modified their client software to make it appear as if they were doing more work than was actually occurring. These users placed a high utility on their ranking in a leader board that recorded the “computation units contributed” for the Seti@Home project. The scoring system did not prevent these rational players from increasing their utility by modifying the behavior of their software .
The things people will do to “be on top” never cease to amaze me.
Alternatively, existing systems may work because there are enough “rational” users that maximize their expected utility by the enjoyment of providing a common good. This altruistic behavior is outside of typical game-theoretic models.
We continue to need better models. Perhaps we should apply the intuitive method to devise some?
Mechanism design can be thought of as inverse game theory – where game theory reasons about how agents will play a game, MD reasons about how to design games that produce desired outcomes.
Ah, so in that respect (and they go on about MD for a while) the second goal was met. I read their three open problems and they didn’t seem to exciting to me, but they did not seem to want to rely on cryptographic techniques as they wanted to do it purely in a AMD/DAMD context which I suppose is interesting research but I’m more than happy to use cryptography.
Ok, and now on to
“Brian F. Cooper and Hector Garcia-Molina, Peer-to-peer resource trading in a reliable distributed system, The 1st International Workshop on Peer-to-Peer Systems (IPTPS’02)”
It looks boring.
It does not seem to assume the self-interest of peers - or at best, it assumes enlightened self-interest. What if peers agree to hold on to some data and then drop it?
I’m glad that in the related work they realized some of the similarities with file systems - for their particular work. Can you imagine if you had to worry about your hard drive being self-interested with respect to what data it was actually going to store or not?
That is it for now.