应用随机进程代数的网络系统可靠性预计方法  

Reliability Prediction for Network Systems Using Stochastic Process Algebra

作　　者：严博[1] 吴晓平[1] 付钰[1] 

机构地区：[1]海军工程大学信息安全系,武汉430033

出　　处：《西安交通大学学报》2011年第6期40-45,共6页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目(60774029)

摘　　要：针对复杂结构网络系统可靠性预计建模与分析的难题,提出了一种应用随机进程代数的网络系统可靠性预计方法.通过分析3种基本网络拓扑结构,研究了网络系统随机进程代数模型的建立方法,根据随机进程代数操作语义,推导出一个带时间延迟的可靠性标记变迁系统,依据其隐含的马尔可夫转移关系和系统的当前状态,计算得到系统在一段时间内的可靠性预计值.仿真实验表明,该方法能精确刻画网络系统随机行为以及组件之间的相互关系,可有效预测不同状态下系统的可靠性水平.A reliability prediction method is proposed based on stochastic process algebra （SPA） to deal with the difficulties in the reliability modeling and analysis of network systems with complex structure. SPA modeling of network systems is proposed through analyzing characters of three basic network topologies. Then a labeled transition system （LTS） with time delay marks for the model is generated utilizing the SPA＇s operational semantics. The prediction value of the system＇s reliability can be solved out from the LTC＇s Markovian transition relation and the system＇s current state. Simulation results show that the proposed method can precisely describe system＇s stochastic behaviors and interactions among components, and predict the reliability level under different status effectively.

关 键 词：网络系统 拓扑结构 随机进程代数 可靠性预计 

分 类 号：TP302.7[自动化与计算机技术—计算机系统结构]