Method for modeling and analysis real-time system dependability using AADL

As real-time systems are becoming more and more complicated, how to analyze and evaluate system dependability indices at the early stage of system development, such as reliability, availability and safety, is a crucial means of guaranteeing the usability of the developed systems. Dealing with this problem, a method of modeling and analyzing the system dependability using AADL (Architecture Analysis and Design Language) was proposed. Firstly, the system AADL dependability model was established by using both the AADL architecture model and the error model. Secondly, the system AADL dependability was transformed to Stochastic Petri Nets and Markov Chains, and the transformation rules among the models were designed. Finally, the dependability indices were calculated by MC method and a simulation was presented to validate the feasibility of the proposed method.