基于三方博弈的P2P流量优化研究

发布时间：2018-04-08 22:18

本文选题：对等网络　切入点：流量优化　出处：《湖北工业大学》2014年硕士论文

【摘要】：随着互联网的发展，，对等网络技术（Peer-to-Peer,P2P）在文件下载、流媒体、即时消息以及VoIP等领域得到了广泛应用。P2P应用在不断丰富网络应用的同时，也消耗网络50%～90%的网络带宽，对Internet服务提供商（Internet ServiceProvider, ISP）的冲击很大，造成了网络拥塞，信息安全等问题。因此，优化P2P的流量成为解决P2P内容提供商（Peer-to-Peer Content Provider, PCP）和ISP之间紧张关系的重要途径。 P2P流量优化本质是减少域间流量传输，主要思想是流量本地化，核心是促进P2P与ISP之间的合作。本文围绕基于三方博弈的P2P流量优化框架，对基于缓存博弈的P2P流量优化模型和基于合作博弈的P2P流量优化模型进行了深入研究，主要工作如下：（1）基于三方博弈的P2P流量优化框架。在对从基于PCP的逆向工程传输优化技术到基于ISP主导的P2P流量管理技术的分析基础上，本文认为ISP、PCP和P2P三方博弈管理是P2P流量优化管理的趋势，并提出了基于三方博弈的P2P流量优化框架，并细化成两个具体的P2P流量优化模型。（2）基于缓存博弈的P2P流量优化模型。在P2P网络中，逻辑网络拓扑和实际的网络拓扑的不匹配是导致大量域间P2P流量产生的重要原因，基于部署缓存的流量本地化思想能很好的解决这一问题。在对缓存部署机制分析的基础上，提出了包括Tracker服务器、ISP服务器、Cache服务器和P2P用户节点的缓存博弈模型，将流量优化归结为以缓存总收益作为优化目标的优化问题，提出了一种基于分段缓存思想的缓存策略。仿真实验表明，缓存博弈能明显地减少域间流量传输，并使域间传输成本过高的ISP获得正收益。（3）基于合作博弈的P2P流量负载均衡模型。在对基于博弈论的P2P流量优化研究进行归纳总结的基础上，从流量负载均衡着手对P2P流量进行优化，提出了基于合作博弈的P2P流量负载均衡模型，并推导求得模型的均衡解，从演化迭代角度对均衡的动态收敛性进行了讨论，然后提出了流量负载均衡算法（TLBA）并从公平性角度对算法进行了分析。最后的仿真实验表明，基于合作博弈的P2P流量负载均衡模型能较好的实现ISP均衡的处理P2P流量请求。
[Abstract]:With the development of Internet, Peer-to-PeerP (P2P) technology has been widely used in the fields of file downloading, streaming media, instant messaging and VoIP.The impact on Internet Service provider (ISP) is very great, which causes network congestion, information security and so on.Therefore, optimizing P2P traffic becomes an important way to solve the tension between Peer-to-Peer Content provider and ISP.P2P traffic optimization is essentially to reduce inter-domain traffic transmission, the main idea is traffic localization, the core is to promote the cooperation between P2P and ISP.In this paper, the P2P traffic optimization model based on cache game and the P2P traffic optimization model based on cooperative game are deeply studied around the framework of P2P traffic optimization based on tripartite game. The main work is as follows:1) P2P traffic optimization framework based on tripartite game.Based on the analysis of the reverse engineering transmission optimization technology based on PCP and P2P traffic management technology based on ISP, this paper thinks that the game management between ISP and P2P is the trend of P2P traffic optimization management.A P2P traffic optimization framework based on tripartite game is proposed and refined into two specific P2P traffic optimization models.2) P2P traffic optimization model based on cache game.In P2P networks, the mismatch between logical network topology and actual network topology is an important reason that leads to a large number of inter-domain P2P traffic. The idea of traffic localization based on deployment cache can solve this problem very well.Based on the analysis of cache deployment mechanism, a cache game model including Tracker server / cache server and P2P user node is proposed. The optimization of traffic is attributed to the optimization of total revenue of cache.This paper presents a caching strategy based on segmented caching.Simulation results show that cache game can significantly reduce inter-domain traffic transmission and make the ISP with high inter-domain transmission cost gain positive income.3) P2P traffic load balancing model based on cooperative game.On the basis of summing up the P2P traffic optimization research based on game theory, this paper optimizes the P2P traffic from traffic load balance, proposes a P2P traffic load balancing model based on cooperative game, and deduces the equilibrium solution of the model.The dynamic convergence of equilibrium is discussed from the point of view of evolutionary iteration, and then the traffic load balancing algorithm (TLBA) is proposed and analyzed from the point of view of fairness.Finally, the simulation results show that the P2P traffic load balancing model based on cooperative game can achieve ISP balance in handling P2P traffic requests.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.02

【参考文献】