基于SVM的网格资源信誉评价

SVM-based reputation evaluation for grid resources

网格的开放性及动态性带来了便利的同时也带来了更多的安全挑战。针对网格这个特定环境,提出了资源调度过程中的信誉评价问题,并给出了相关的信誉存储模式。在分析网格成员特点及SVM原理的基础上提出了基于SVM的信誉评价算法,并对该算法进行了仿真实验。仿真实验结果表明基于SVM的信誉评价算法能够较好地适应网格环境并能较客观及准确地评价网格成员的信誉。

The open and dynamic grid brings not only facilities but also great security challenges. The reputation evaluation issue of resource scheduling is proposed in the special grid environment. A relevant reputation memory model is also put forward. SVM-based reputation evaluation algorithm is presented after the analysis of characteristics of grid members and the principle of SVM. Its simulation experiment is also done. The experimental results show that SVM-based reputation evaluation algorithm can adapt to the grid environment and describe the reputation of grid members well.