使用兴趣子网划分算法对Gnutella中资源定位机制的改进

Gnutella是一种对等网络文件共享应用，使用“洪泛”算法进行资源定位，具有简单、容错的特点，但扩展性差．提出一种基于兴趣本体模型对用户兴趣进行建模，通过将具有相似兴趣的节点建立直接连接构成兴趣子网，从而有效地避免在Gnutella中使用广播扩散方式进行资源定位，以提高Gnutella的可扩展性和内容定位性能．仿真实验证明了兴趣子网模型是有效的。     

基于任务复制的进化调度算法

并行程序任务调度是NP完全问题。已有的进化调度算法不能很好利用任务复制这一重要手段。为此，利用染色体直接编码和间接编码的优点，并结合任务图的知识，使进化调度算法的搜索能力得到增强，同时染色体的解码过程也得到简化。通过与同类算法的性能比较，表明了本算法在调度问题中的有效性。     

基于凸多面体剖分的并行碰撞检测算法

针对传统的碰撞检测算法对于复杂多面体之间的碰撞检测时间过长的缺陷，提出并证明了一种凸多面体的剖分算法，该算法能将复杂的凸多面体剖分为简单四面体，利用基于MPI(Message Passing Interface)并行处理的方法来提高碰撞检测的实时性．给出了基于凸多面体剖分的并行碰撞检测算法与常规算法的比较结果，证明了所提出算法的优越性．     

Improvement of Incentive Mechanism on BitTorrent-like Peer-to-Peer File Sharing Systems

BitTorrent is a very popular Peer.to-Peer file sharing system, which adopts a set of incentive mechanisms to encourage contribution and prevent free-riding. However,we find that BitTorrent's incentive mechanism can prevent free-riding effectively in a system with a relatively low number of seeds, but may fail in producing a disincentive for free-riding in a system with a high number of seeds. The reason is that BitTorrent does not provide effective mechanisms for seeds to guard against free-riding.Therefore, we propose a seed bandwidth allocation strategy for the BitTorrent system to reduce the effect of seeds on free-riding. Our target is that a downloader which provides more service to the system will be granted a higher benefit than downloaders which provide lower service when some downioaders ask for downloading file from a seed. Finally,simulation results are given, which validate the effectiveness of the proposed strategy.     

基于P2P的通告代理集群技术

提出基于P2P网络技术构建Web服务通告代理集群,同时,提出对通告代理集群进行基于主题的逻辑多层管理,在主题层内使用层内通告消息广播算法传播通告消息.仿真结果表明,通告代理集群对随机节点故障是强壮的,基于主题的分层技术有效减少了每个通告代理节点的负载,层内通告消息广播算法具有高可靠性和好的可扩展能力.