|
CMU-CS-02-164
Computer Science Department
School of Computer Science, Carnegie Mellon University
CMU-CS-02-164
Profit Maximizing Mechanisms for the
Extended Multicasting Game
Shuchi Chawla, David Kitchin, Uday Rajan*, Ramamoorthi Ravi*, Amitabh Sinha
July 2002
CMU-CS-02-164.ps
CMU-CS-02-164.pdf
Keywords:Game theory, mechanism design, approximation
algorithms, multicasting game
We consider the design of multicast networks when both edges and nodes are
selfish agents. Our objective is a budget-balanced, strategy-proof mechanism
which selects the set of clients to receive service and constructs a
network to provide the service. It extracts payments from the
clients and pays edges to participate in the network, and aims to
maximize profit from the transaction. We introduce the notion of
@i(profit guaranteeing) mechanisms, and show the existence of one such
mechanism. The mechanism provides guarantees of the form that in a
sufficiently profitable market, it obtains some fraction of the obtainable
profit, and if the market is sufficiently unprofitable, then the mechanism
demonstrates that no profitable solution exists. The
mechanism runs in polynomial time. To our knowledge, this is the
first study of mechanisms for designing multicast networks in which
edge values are unknown.
11 pages
*Graduate School of Industrial Administration, Carnegie Mellon University
|