计算机工程

• •    

基于AADL的IoP系统可靠性评估方法

  

  • 发布日期:2021-01-08

An AADL-based dependability evaluation method for IoP System

  • Published:2021-01-08

摘要: 人联网(IoP)是一种以人为中心的新型实时交互式物联网,基于 IoP 系统的可靠性分析是最新研究领域之一。由 于 IoP 系统中包含了复杂的架构和海量、实时变化的数据,使得基于 IoP 系统的可靠性分析变得十分困难。目前,该领域依 然缺乏一种健全的基于 IoP 系统的可靠性建模及评估方法。针对上述问题,本文提出了一种基于架构分析与设计语言(AADL) 的 IoP 系统可靠性评估方法,该方法首先利用 AADL 及其附件语言对 IoP 系统进行可靠性建模,接着基于可靠性模型从定性 角度评估了系统故障的根本原因和风险,然后 结合 Ocarina 模型转换技术提出了一种基于连续时间马尔科夫链(CTMC)的定 量评估算法,该算法将 AADL 可靠性模型转换为 CTMC 模型,实现了 对 系统动态和实时等特性的评估。最后设计了一个 IoP 系统通用模型,并以此为案例验证了所提出方法的可行性。该方法不仅能对 IoP 系统进行完善地建模,还能自动、准确地对 其进行可靠性分析,具有良好的应用价值。

Abstract: 】 The Internet of People (IoP) is a new human-centric and real-time interactive Internet of Things. The dependability analysis for IoP system is one of the up-to-date research areas in recent years. IoP system includes complex architecture and a massive volume of real-time changing data. These cause the dependability analysis based on IoP system much more complicated. Currently, there is still a lack of a robust dependability modelling and analysis method for IoP system. To solve the problem, an AADL-based dependability evaluation method for IoP system is proposed. Firstly, the method uses AADL and its annex language to build IoP system's dependability model, then based on the dependability model, the failure cause and its hazards of IoP system are evaluated qualitatively. Furthermore, by combining the Ocarina model transformation technology, a quantitative evaluation algorithm based on the Continuous-time Markov Chain (CTMC) is proposed. The algorithm transforms the AADL dependability model to the CTMC model, so that the dynamic and real-time attributes of IoP system can be evaluated quantitatively. Finally, a general IoP system model is designed, which is used to demonstrate the feasibility of the proposed method. The proposed evaluation method can not only model IoP system but also can automatically and accurately analyze the dependability of IoP system, which has a good application value.