Distributed Algorithmic Mechanism Design (DAMD) combines theoretical computer science's traditional focus on computational tractability with its more recent interest in incentive compatibility and distributed computing. The Internet's decentralized nature, in which distributed computation and autonomous agents prevail, makes DAMD a very natural approach for many Internet problems. This paper first outlines the basics of DAMD and then reviews previous DAMD results on multicast cost sharing and interdomain routing. The remainder of the paper describes several promising research directions and poses some specific open problems.

Abstract not found

In the efficient social choice problem, the goal is to assign values, subject to side constraints, to a set of variables to maximize the total utility across a population of agents, where each agent has private information about its utility function. In this paper we model the social choice problem as a distributed constraint optimization problem (DCOP), in which each agent can communicate with other agents that share an interest in one or more variables. Whereas existing DCOP algorithms can be easily manipulated by an agent, either by misreporting private information or deviating from the algorithm, we introduce M-DPOP, the first DCOP algorithm that provides a faithful distributed implementation for efficient social choice. This provides a concrete example of how the methods of mechanism design can be unified with those of distributed optimization. Faithfulness ensures that no agent can benefit by unilaterally deviating from any aspect of the protocol, neither informationrevelation, computation, nor communication, and whatever the private information of other agents. We allow for payments by agents to a central bank, which is the only central authority that we require. To achieve faithfulness, we carefully integrate the Vickrey-Clarke-Groves (VCG) mechanism with the DPOP algorithm, such that each agent is only asked to perform computation, report

Mechanism design (MD) provides a useful method to implement outcomes with desirable properties in systems with self-interested computational agents. One drawback, however, is that computation is implicitly centralized in MD theory, with a central planner taking all decisions. We consider distributed implementations, in which the outcome is determined by the self-interested agents themselves. Clearly this introduces new opportunities for manipulation. We propose a number of principles to guide the distribution of computation, focusing in particular on Vickrey-Clarke-Groves mechanisms for implementing outcomes that maximize total value across agents. Our solutions bring the complete implementation into an ex post Nash equilibrium.

Secure Computation essentially guarantees that whatever computation n players can do with the help of a trusted party, they can also do by themselves. Fundamentally, however, this notion depends on the honesty of at least some players. We put forward and implement a stronger notion, Rational Secure Computation, that does not depend on player honesty, but solely on player rationality. The key to our implementation is showing that the ballotbox—the venerable device used throughout the world to tally secret votes securely—can actually be used to securely compute any function. Our work bridges the fields of Game Theory and Cryptography, and has broad implications for Mechanism Design. 1 The Case for Rational Security Secure Computation. The general notion of Secure Computation was put forward and first exemplified by Goldreich, Micali and Wigderson [8], building on earlier two-party results of Yao [17]. Given a joint computation among n players and a trusted party, Secure Computation aims at removing the trusted party without suffering any correctness or privacy loss. A bit more precisely, all prior secure-computation work— by now quite extensive — adopts the original ideal/real paradigm, illustrated below in the crucial, special case of a secure function evaluation (SFE for short). An ideal evaluation of a (possibly probabilistic) ninput, n-output function f consists of the following process. Each player i has a private input, xi, and is assumed to be honest or malicious. An honest i simply confides his original xi to a trusted party. Malicious players may instead perfectly coordinate their actions, so as to compute and report to the trusted party alternative inputs x ′ j for every malicious player j. The trusted party then evaluates f on all reported inputs,

We study mechanisms for utilitarian combinatorial allocation problems, where agents are not assumed to be singleminded. This class of problems includes combinatorial auctions, multi-unit auctions, unsplittable flow problems, and others. We focus on the problem of designing mechanisms that approximately optimize social welfare at every Bayes-Nash equilibrium (BNE), which is the standard notion of equilibrium in settings of incomplete information. For a broad class of greedy approximation algorithms, we give a general black-box reduction to deterministic mechanisms with almost no loss to the approximation ratio at any BNE. We also consider the special case of Nash equilibria in fullinformation games, where we obtain tightened results. This solution concept is closely related to the well-studied price of anarchy. Furthermore, for a rich subclass of allocation problems, pure Nash equilibria are guaranteed to exist for our mechanisms. For many problems, the approximation factors we obtain at equilibrium improve upon the best known results for deterministic truthful mechanisms. In particular, we exhibit a simple deterministic mechanism for general combinatorial auctions that obtains an O ( √ m) approximation at every BNE. 1

Distributed algorithmic mechanism design (DAMD) is an approach to designing distributed systems that takes into account both the distributed-computational environment and the incentives of autonomous agents. In this dissertation, we study two problems, multicast cost sharing and interdomain routing. We also touch upon several issues important to DAMD in general, including approximation, compatibility with existing protocols, and hardness that results from the interplay of incentives and distributed computation.

We continue the study of multicast cost...

We characterize the outcomes of games when players may make binding offers of strategy contingent side payments before the game is played. This does not always lead to efficient outcomes, despite complete information and costless contracting. The characterizations are illustrated in a series of examples, including voluntary contribution public good games, Cournot and Bertrand oligopoly, principal-agent problems, and commons games, among others.

We give a brief and biased survey of the past, present, and future of research on the interface of theoretical computer science and game theory. 1