Essays in peer-to-peer content distribution networks: Scale, formation, and pricing.

摘要

Peer-to-Peer (P2P) networks (Gram 2001) are emerging as increasingly important distribution channels for digital goods. This dissertation includes three research topics, each of which separately addresses the operational (scale), social (formation), and economic (pricing) issues of P2P networks and provides guidelines for respective decision making.; We first study both the positive and negative scale effects on the operations of P2P file sharing networks, and propose optimal sizing (optimal number of peers) and grouping (optimal number of directory intermediary) decisions. Our results show that increasing network scale has a positive effect on the expected content availability and the transmission cost, but a negative effect on the expected provision and the search costs. Furthermore, we extend the model to examine the grouping decision in the networks with symmetric interconnection structures, and compare the performance between random- and location-based grouping polices.; We further utilize theories of social networks and economic incentives to investigate the formation of P2P networks with rational participating agents. A framework for multi-level formation dynamics, including an individual level (content-sharing decision and group selection) and a group level (membership admission and interconnection), is proposed to examine the emergence of self-organized P2P network configuration features such as free-riding level, group size, and link structure. Furthermore, we investigate the stability and efficiency of P2P networks, and propose internal transfer mechanisms to force stable networks to become efficient.; Finally, we develop the pricing schemes for services, such as content provision, search and transfer, in P2P networks. Consistent with usual implications of public goods provision, our model shows that, in the absence of appropriate incentive mechanisms, P2P users tend to over-request and under-provide contents. To regain the inherent efficiency loss due to decentralized decision-making, we propose a pricing scheme that charges request fees for search intermediary service and compensates content provision so as to induce socially optimal content request and provision decisions by individuals. We also investigate the pricing schemes from the perspectives of a monopolistic profit-seeking P2P provider as well as a provider in a competitive market.