The economic model of network services, both services for basic connectivity as well as application layer services delivered via the network, is undergoing a period of unprecedented change. For instance, Internet Service Providers (ISPs) that offer services such as voice over IP are deciding to use proprietary forms of service differentiation to improve these service offerings over those of competitors. As more ISPs implement service differentiation in a proprietary way, there is a danger that the opportunity to have a coherent architectural approach to service differentiation will be lost. This issue is tied to the net neutrality debate, in which one core issue is how ISPs and content/application service providers should share revenue. Some fundamental questions include the following. What are the market effects of having only competition between proprietary service stacks versus having competition at each layer? What architectural principles should be established in law or standards to induce the development of a market structure that results in higher welfare for all the stakeholders? Should both standard interfaces and/or charging practices be enforced between providers? This project investigates questions such as these in order to achieve a more complete understanding of the coupling of architecture and economics in network services. The project develops game theoretic models of the interaction of selfish users and providers, and then uses these models to understand the welfare effects of different architectural choices. The results of this project will inform the development of both the architecture and market structure of the network services industry.
