Game theory for collaboration in future networks

Abstract

Cooperative strategies have the great potential of improving network performance and spectrum utilization in future networking environments. This new paradigm in terms of network management, however, requires a novel design and analysis framework targeting a highly flexible networking solution with a distributed architecture. Game Theory is very suitable for this task, since it is a comprehensive mathematical tool for modeling the highly complex interactions among distributed and intelligent decision makers. In this way, the more convenient management policies for the diverse players (e.g. content providers, cloud providers, home providers, brokers, network providers or users) should be found to optimize the performance of the overall network infrastructure. The authors discuss in this chapter several Game Theory models/concepts that are highly relevant for enabling collaboration among the diverse players, using different ways to incentivize it, namely through pricing or reputation. In addition, the authors highlight several related open problems, such as the lack of proper models for dynamic and incomplete information games in this area.