基于测量的p2p VoD系统的模型研究

peer-to-peer方式的VoD系统中,用户的观看行为、网络中资源块(chunk)的分布、下载速度和pre-fetch机制之间相互影响,相互制约。以往的研究通常专注于某个方面,而没有考虑各因素间的相互关系。本文通过对实际数据进行分析,建立模型描述p2p VoD系统中上述各因素间的关系,并基于该模型提出一种新的pre-fetch机制,能在保证用户流畅观看的前提下,尽可能提高下载速度。该模型提供一种更为精确的了解客户端行为的方法,为其他各类p2p应用中研究用户行为、设计下载机制、优化系统提供参考。     

A secure trust model for P2P systems based on trusted computing

Trust is one of the most important security requirements in the design and implementation of peer-to-peer （P2P） systems. In an environment where peers＇ identity privacy is important, it may conflict with trustworthiness that is based on the knowledge related to the peer＇s identity, while identity privacy is usually achieved by hiding such knowledge. A trust model based on trusted computing （TC） technology was proposed to enhance the identity privacy of peers during the trustworthiness evaluation process between peers from different groups. The simulation results show that, the model can be implemented in an efficient way, and when the degree of anonymity within group （DAWG） is up to 0.6 and the percentage of malicious peers is up to 70%7 the service selection failure rate is less than 0.15.     

PBRRI： A Hybrid Peer-to-Peer Communication Model in Distributed Systems

In distributed systems independent agents need to interact with each other to accomplish their task. Modern peer-to-peer computing technologies always concern with enabling interaction among agents and help them cooperate with each other. But in fact, access control should also be considered to limit interaction to make it harmless. This paper proposed a proxy based rule regulated interaction (PBRRI) model. Role based access control is introduced for security concerns. Regulation rules are enforced in a distributed manner so that PBRRI can be applied to the open distributed systems such as Internet.     

An ID-Based Authenticated Key Agreement Protocol for Peer-to-Peer Computing

Introduction Overrecentyears,withtheemergenceofdis-tributedapplicationsovertheInternet,anewmod-elofcommunicationandcomputation,calledpeer-to-peer(P2P)computing,hasgainedgreatpopu-larity[1-3].Contrarytothetraditionalclient-server model,P2Pcomputingeliminatesthenotionof“server”,andallpeersinP2Psystemcanactas clientsandserversatthesametime.Therefore,thesinglepointoffailureandthebottleneckof“server”encounteredinthetraditionalclient-serv-ermodelcanbeeliminated,andtheperformance andreliabilit…     

基于NDIS中间层驱动的对等网络流量监管机制

由于对等网络P2P对网络带宽的贪婪性,对P2P数据流的监管是当前研究的热点问题.本文在分析比较几种相关技术的基础上,提出了一个基于网络驱动接口规范（NetworkDriver Interface Specification,NDIS）中间层驱动的对等网络监管系统架构,对各功能实体特性进行了详细描述.以设备驱动程序开发包（Device Driver Kit,DDK）提供的Passthru例程为骨架对该系统进行了实现,对共享内存通信机制、数据拦截机制和主要数据结构进行了讨论.最后,通过测试数据拦截完备性和效率,以及对P2P软件的识别和阻断能力,验证了本系统的有效性和可用性.     

Low cost and reliable anonymity protocols in P2P reputation systems

In peer-to-peer (P2P) reputation systems,each peer's trustworthiness is evaluated based on its pseudonym's rating values given by other peers. Since it is assumed that each peer has a long lived pseudonym,all the transactions conducted by the same peer may be linked by its pseudonym. Therefore,one of the fundamental challenges in P2P reputation systems is to protect peers' identity privacy. In this paper,we present two independent anonymity protocols to achieve all the peers' anonymity by changing pseudonym...     

基于遗传算法的P2P激励机制

针对P2P系统中共享存储资源的搭便车和公共悲剧问题，提出了一种基于遗传算法的P2P存储资源共享激励机制，提供资源的节点根据遗传算法选择最优策略分配资源，以使其贡献值最大化，贡献值大的节点其奇支配使用的共享存储资源越多，仿真实验表明，该机制能有效地实现P2P系统中存储资源分配的公平性和效率，达到了激励节点参与共享资源，抑制节点自私行为的目的．     

一种具有递归结构和均匀度分布的P2P协议

针对DHT类型的P2P协议在网络演进的过程中会产生超级节点,进而导致度分布严重失衡,网络传输的平均时延增大的问题,该文在Chord协议的基础上提出了一种新的具有递归结构的P2P协议：R—Chord,并进行了仿真分析．结果说明,该协议能够有效抑制超级节点的产生,使节点的度分布趋于均匀,使网络的负载在节点间的分布更均衡．     

Caching Scheme Based on Power-Law Characteristic in Unstructured P2P Networks

In order to reduce the traffic load and improve the availability of the shared resources in unstructured P2P networks, a caching scheme combining alternative index and adaptive replication (AIAR) is presented. AIAR uses random walk mechanism to disperse the caching information of resources in the network based on its power-law characteristic, and dynamically adjusts replicas according to the visit frequency on resources and the degree information of peers. Subsequent experimental results show that the proposed AIAR scheme is beneficial to improve the search performance of success rate and respond speed. In addition, compared to some existing caching scheme, AIAR can perform much better in success rate, especially in a dynamic environment.     

基于马尔可夫决策的P2P激励机制研究

对网络节点状态进行分类,找寻网络节点状态的转换趋势,针对节点自私化的这一趋势,在网络中设置一种基于马尔可夫链的预测机制,根据节点合作率地高低,进行相应的约束措施.并针对网络节点博弈现状对节点未来状态进行预测,以此来做出更为及时地决策,利用Matlab软件对预测机制对网络的影响进行了仿真与对比,证实了该机制在宏观上对于网络的稳定性与节点的合作度具有有效地激励作用.