Costas Courcoubetis, Incentives for Large Peer-to-Peer Systems, May 17th, 2006 Abstract: In economics, our problem is known as the Mechanism Design problem. We need to elicit truthful information from the agents, or peers, regarding their valuation of the service, set Q, and decide which of them are allowed to participate in using the system and how much each should contribute to covering the cost of building the system at level Q. This is to be done to produce the greatest possible social welfare. While the full solution of this problem is extremely complex and not easily solved in practice, we show that as the number of agents becomes large, there is a good solution to our problem that takes a very simple form. We merely require each agent to pay the same fixed fee towards payment of the total cost, and exclude agents that are unwilling to do so. In the cases we consider, this fee need not be paid in cash, but can be paid in kind, i.e., by contributing to a fixed part of the overall service. Such a simple contribution policy is easy to implement and requires no centralized implementation. The only information required by the system designer to compute the fixed fee is the distribution of the agents' valuations for the service. We also discuss extensions of the model where peers can choose certain parameters that may affect their behaviour and suggest some possible applications. Our first application is to a model of file sharing, in which the public good is content availability; the second concerns a problem of peering wireless LANs, in which the public good is the availability of connectivity for roaming peers. In both problems we can cope with the requirement that the payments be made in kind, rather than in cash.
CV: His current research interests are economics of networks with emphasis in the development of pricing schemes that reduce congestion and enhance stability and robustness, quality of service and management of integrated services, performance and traffic analysis of large systems, applied probability models. Other interests include the combination of e-commerce technologies with telecommunications, and formal methods for software verification. He has published over 80 papers in scientific journals such as Operations Research, Mathematics of Operations Research, Journal on Applied Probability, IEEE Transactions in Communications, IEEE Transactions in Automatic Control, IEEE JSAC, SIAM Journal on Computing, Stochastic Processes, Probability in Engineering and Information Sciences, Queuing Systems, ACM TOPLAS, JACM, Formal Methods in System Design, Telecommunications Systems, Information and Computation, Theoretical Computer Science, and in conferences such as FOCS, STOC, LICS, INFOCOM. GLOBCOM, ITC, ACM SIGMETRICS. His work has over 2400 citations according to the NECI Scientific Literature Digital Library. He is a co-author with Richard Weber of “Pricing Communication Networks: Economics, Technology and Modeling” (Wiley, 2003). |