Internet economics and ISP business settlements.

摘要

Internet service providers (ISPs) depend on one another to provide global network services. Within the current Internet, autonomous ISPs implement bilateral settlements, with each ISP establishing contracts that suit its own local objective to maximize its profit. However, the profit-seeking nature of the ISPs leads to selfish behaviors that result in inefficiencies and disputes in the network.;From a macroscopic view, this concern is at the heart of the network neutrality debate, which argues whether content-based service differentiation should be allowed on the Internet. It asks for an appropriate compensation structure to resolve ISP disputes as well as to provide an efficient and well-connected network. From a microscopic view, this concern manifests in ISP selfish routing strategies and discriminatory interconnections, which limit the stability of routes, discourage potentially useful connectivity and deteriorate performance and profit of the network.;In this dissertation, we study the use of Shapley value, which is originated from coalition game theory, as a profit-sharing mechanism to make ISP settlements. From a macroscopic view, we model a detailed and realistic network with three classes of ISPs: content, transit, and eyeball, as well as two types of user demand: elastic and inelastic. We derive closed-form Shapley values for structured ISP topologies and develop a dynamic programming procedure to compute the Shapley values under more diverse Internet topologies. From a microscopic view, we model a generic network with three layers of ISP decisions: interconnecting, routing and financial settlement. We re-design the current bilateral financial settlements by a clean-slate multi-lateral profit-sharing mechanism based on the Shapley value. We show that if the Shapley profit-distribution is enforced at a global level, then ISPs' selfish interconnecting and routing strategies will converge to a Nash equilibrium, at which these individual strategies maximize the aggregate network profit and encourage ISP connectivity so as to limit Internet balkanization. From a practical point of view, we explore our results' implication on the appropriate bilateral settlements between ISPs and the pricing structures for differentiated services. We conclude that our Shapley framework provides an ideal guideline and/or benchmark for solving disputes between ISPs and for building regulatory protocols for the Internet industry.