PMC模型下的h-边容错诊断数  

h-edge tolerant diagnosabilities under the PMC model

作　　者：张念蓬 朱强 ZHANG Nianpeng;ZHU Qiang(School of Mathematics and Statistics, Xidian University, Xi′an 710071, China)

机构地区：[1]西安电子科技大学数学与统计学院,陕西西安710071

出　　处：《西北大学学报（自然科学版）》2020年第5期787-792,共6页Journal of Northwest University（Natural Science Edition）

基　　金：国家自然科学基金资助项目(61672025)。

摘　　要：系统级诊断是多处理器系统设计和维护中的重要方面,通过诊断参数来衡量系统的容错性能。传统的诊断参数都是在假设系统中仅有处理器发生故障的情形下得到的,但是在实际情形中,系统中的处理器和链接都可能发生故障。该文研究了新的系统级诊断参数—h-边容错诊断数。当系统G中的故障边数不超过h时,G中包含的可以被全部识别的最大故障点数称为系统G的h-边容错诊断数。通过对一般图中公共邻点数的限制,证明了PMC模型下一般图的h边容错诊断数。文中确定了k-元n-方体、平衡立方体、交换立方体和交换折叠立方体4类网络在PMC模型下的h-边容错诊断数,为衡量系统在点边混合故障情形下的容错性能提供了有效参数。System-level diagnosis is an important aspect in the design and maintenance of multiprocessor systems,the fault tolerance performance of multiprocessor system is measured by diagnostic parameters.Traditional diagnostic parameters are obtained under the assumption that only the processor in the system has failed.However,in the actual situation,both processors and links in the system may fail.In this paper,a new system-level diagnosis parameter,h-edge tolerant diagnosabilities is studied.When the number of fault edges in the system G does not exceed h,the maximum number of fault points included in the system G that can be guaranteed to be identified is called the system′s h-edge tolerant diagnosabilities.By limiting the common adjacent points in general graphs,the number of h-edge tolerant diagnosabilities in PMC models is proved.In this paper,the h-edge fault tolerant diagnosabilities for the following four types of networks is determined:k-ary n-cubes,balanced hypercubes,exchange hypercubes and exchange folded hypercubes under the PMC model,which provides an effective parameter for measuring the fault-tolerant performance of multiprocessor systems in the case of mixed faults.

关 键 词：多处理器系统 PMC模型 混合故障诊断 h-边容错诊断数 

分 类 号：O614[理学—无机化学]