RE: returning results for partial downloads

On Saturday, October 06, 2007 Michael Rogers wrote:
> I've been working on an extension of the prisoner's dilemma that
> incorporates resource scarcity:
>
> http://www.cs.ucl.ac.uk/staff/mrogers/sharers-dilemma.html

	Nice to see someone doing some math on that subject. However,
after the first cursory reading I was not able to understand - what
is it exactly  that you are aiming to maximize? Is it your goal to 
maximize the cumulative upload rate of the network, or you also 
want to divide it between the downloaded in some fair way? What
is your definition of fair, if this is the case? I mean, you need
to calculate a single one-dimensional value (call it the algorithm
quality or utility) for each algorithm to compare them, right? What 
is this value and how is it calculated?

	Best wishes -
	Oso.

P.S. Zlatin: +1. Long overdue.


-----Original Message-----
From: [email protected] [mailto:[email protected]]On Behalf
Of Michael Rogers
Sent: Saturday, October 06, 2007 4:16 AM
To: [email protected]
Subject: Re: [the_gdf] returning results for partial downloads


Zlatin Balevsky wrote:
> I believe most of these ideas are already implemented in the current
> download mesh.  What is missing are the ideas for queueing discipline.
>  That is a very interesting topic, and it would be a very exciting
> challenge to come up with schemes that can beat bittorrent's
> tit-for-tat.

I thought BitTyrant's approach was interesting: it tries to determine
the threshold at which it will be unchoked by each peer, sorts the peers
by their download/upload ratios at the threshold, and uploads to as many
as possible given the available upstream bandwidth.

http://bittyrant.cs.washington.edu/

In a sense though, it's parasitic on BitTorrent's tit for tat strategy:
it exploits the fact that once you persuade a peer to unchoke you,
there's nothing to be gained by further increasing your upload speed.

It seems to me that the prisoner's dilemma isn't the appropriate model
for this problem: uploading to one neighbour prevents you from uploading
to another, so you can't treat the situation as a set of separate
two-player games. I've been working on an extension of the prisoner's
dilemma that incorporates resource scarcity:

http://www.cs.ucl.ac.uk/staff/mrogers/sharers-dilemma.html

You can see the parallel between BitTyrant's strategy and the utility
maximisation strategy in the sharer's dilemma. Saleem Bhatti and I were
thinking of running a tournament to find good strategies for the
sharer's dilemma. Any takers? :-)

Cheers,
Michael


 
Yahoo! Groups Links