An Analysis of Incentives for Network Infrastructure Investment Under Different Pricing Strategies

• 成果类型: Article
• 署名作者: Gupta, Alok; Jukic, Boris; Stahl, Dale O.; Whinston, Andrew B.
• 署名单位: University of Minnesota System; University of Minnesota Twin Cities; Clarkson University; University of Texas System; University of Texas Austin; University of Texas System; University of Texas Austin
• 刊物名称: INFORMATION SYSTEMS RESEARCH
• ISSN/ISSBN: 1047-7047
• DOI: 10.1287/isre.1090.0253
• 发表日期: 2011
• 页码: 215-232
• 关键词: computer services

摘要

The Internet is making a significant transition from primarily a network of desktop computers to a network variety of connected information devices such as personal digital assistants and global positioning system-based devices. On the other hand, new paradigms such as overlay networks are defining service-based logical architecture for the network services that make locating content and routing more efficient. Along with Internet2's proposed service-based routing, overlay networks will create a new set of challenges in the provision and management of content over the network. However, a lack of proper infrastructure investment incentive may lead to an environment where network growth may not keep pace with the service requirements. In this paper, we present an analysis of investment incentives for network infrastructure owners under two different pricing strategies: congestion-based negative externality pricing and the prevalent flat-rate pricing. We develop a theoretically motivated gradient-based heuristic to compute maximum capacity that a network provider will be willing to invest in under different pricing schemes. The heuristic appropriates different capacities to different network components based on demand for these components. We then use a simulation model to compare the impact of dynamic congestion-based pricing with flat-rate pricing on the choice of capacity level by the infrastructure provider. The simulation model implements the heuristic and ensures that near-optimal level of capacity is allocated to each network component by checking theoretical optimality conditions. We investigate the impact of a variety of factors, including the per unit cost of capacity of a network resource, average value of the users' requests, average level of users' tolerance for delay, and the level of exogenous demand for services on the network. Our results indicate that relationships between these factors are crucial in determining which of the two pricing schemes results in a higher level of socially optimal network capacity. The simulation results provide a possible explanation for the evolution of the Internet pricing from time-based to flat-rate pricing. The results also indicate that regardless of how these factors are related, the average stream of the net benefits realized under congestion-based pricing tends to be higher than the average net benefits realized under flat-rate pricing. These central results point to the fallacy of the arguments presented by the supporters of net neutrality that do not consider the incentives for private investment in network capacity.