移动P2P流媒体中带QoS保障的契约激励技术研究

Current commercial mobile streaming applications call for more stable and delicate technologies of QoS guarantee. In this project, we provide a comprehensive treatment of incentive mechanisms with QoS guarantee under novel mobile P2P (peer-to-peer) streaming architecture. (1) In particular, we first investigate the incentive mechanism among mobile peers by modeling peers' caching activity as a caching game. The goal of this part of research is to find the principle between pricing and the enhancement of network performance. (2) In what follows, we envision mobile P2P streaming system as a QoS trading market and build a contract-ruled incentive model with QoS guarantee. We study the Nash equilibrium of the incentive model. Considering the free-riding problem, we design corresponding incentive mechanism to derive the agreed QoS without violating self-interested of involved parties. (3) We investigate the performance of Nash equilibrium of the incentive model by introducing the concept of POA(price of anarchy), and propose an optimal pricing strategy through which higher Pareto efficiency can be achieved.. (4) Since assisting peers' remuneration is closely related to the private QoS information reported by peers themselves, we propose a cost-effective and strategy-proof payment algorithm based on Vickrey-Clarke-Groves (VCG) mechanism to encourage the agents to report their costs truthfully. The novelty of this project is to regulate the roaming peers in the context of contracts to fulfill contractual obligations, as well as motivate peers' self-interests. We focus more on economic analysis from macroscopic view, instead of concentrating on the detail technical issues on system implementation. We believe that it could shed light on the deployment and evolution of practical mobile streaming markets.

现有移动P2P环境中的激励方法无法满足移动流媒体应用中持续稳定细化的QoS保障要求。本项目以移动P2P流媒体系统为背景，研究带QoS保障的协同激励技术，包括：(1)移动节点的缓存激励研究，揭示利用价格工具提升网络性能的规律；(2)利用契约理论建立带QoS保障的激励模型，推导博弈模型的均衡状态，理论分析协助节点"搭顺车"的成因，设计并实现激励机制来保障系统获得契约规定的QoS；(3)激励模型中的效率分析及其优化价格策略研究,利用博弈论工具分析激励模型的效率损失，提出价格优化策略以改进全局福利；(4)成本优化的协助节点策略性防伪研究，分析协助节点中可能存在的隐藏信息问题，设计并实现一个成本优化且具有策略性防伪功能的支付算法。本项目研究从网络经济学和博弈论相结合的独特视角，利用契约理论来细致有效地约束各参与实体行为，在满足激励相容的前提下，达到系统QoS保障的目的，具有重要的科学和借鉴意义。

现有移动P2P环境中的激励方法无法满足移动流媒体应用中持续稳定细化的QoS保障要求。本项目以移动P2P流媒体系统为背景，研究了带QoS保障的协同激励技术，具体包括：(1) 研究了移动节点的缓存规律，提出了利用价格工具提升网络性能的方法；(2)利用契约理论建立了带QoS保障的激励模型，推导得出博弈模型的均衡状态，理论分析了协助节点“搭顺车”的成因，设计并实现了激励机制来保障系统获得契约规定的QoS；(3)利用博弈论工具分析了激励模型的效率损失，提出了价格优化策略以改进全局福利；(4)分析了协助节点中可能存在的隐藏信息问题，设计并实现了一个成本优化且具有策略性防伪功能的支付算法。实验和仿真结果表明，利用契约理论可以细致有效地约束各参与实体行为，在满足激励相容的前提下，达到了系统QoS保障的目的。本项目的研究可为移动P2P流媒体系统中服务质量保障及激励技术的研究提供一个新的视角，具有重要的科学意义。