Enforcing Fairness in Asynchronous Collaborative Environments

Abstract

Many large-scale distributed applications rely on collaboration, where unrelated users or organizations share their resources for everyone's benefit. However, in such environments any node may attempt to maximize its own benefit by exploiting other's resources without contributing back. Collaborative systems must therefore deploy strategies to fight free-riders, and enforce collaborative behavior. This chapter explores a family of mechanisms to enforce fairness in asynchronous collaborative environments, where simple tit-for-tat policies cannot be used. Our solutions rely on enforced neighborhood relations, where each node is restricted in the choice of other nodes to collaborate with. This creates long-term collaboration relationships, where each node must behave well with its neighbors if it wants to be able to use their resources.