网格计算系统中的综合信任评估

Integration Trust Evaluation in Grid Computation System

网格是个动态变化的环境,网格站点间的信任关系受到一些因素的制约,该文提出了一个网格安全信任模型,把信任的概念建立在网格站点的身份信任和行为信任、计算能力、单元服务开销、先前的工作成功率、入侵检测和入侵抵抗能力的基础上,构造了一个评估网格站点行为信任度的算法。基于模糊数学中的模糊综合评估算法理论,该文构造了一个网格站点信任度的综合评估算法,利用该算法可以比较全面地评估网格站点的信任度,最大程度地保证网格行为的安全可靠。

Grid computation system is a dynamically changed environment, the trust relationship among sites is restricted by some factors. This paper proposes a new grid security trust model, and establishes the concept of trust on the foundation of grid site＇s identity trustworthiness and behavior trustworthiness, computing power, unit server cust, previous success rate of job, intrusion detection capability and intrusion resistance capability. A new algorithm to evaluate grid site＇s behavior trustworthiness is constructed. An algorithm to evaluate grid site＇s integrated trustworthiness is proposed based on fuzzy integration evaluation theory of fuzzy mathematics. By means of the algorithm, trustworthiness among sites can be evaluated comprehensively and completely. Security and reliability of grid behavior can be guaranteed greatly.