Reliability Analysis Of AFDX Network Based On State/Event Fault Tree

In safety critical domains,such as industrial control,avionics and automotive,software components instead of mechanical and electronic components have become a trend.System security analysis is mainly to determine the key behavior and their combinations which cause the failures,and to determine the system design deficiencies.AFDX network is a data network in avionics systems,with high reliability,resilience against harsh environments and real time.Existing AFDX network reliability modeling is based on Fault Tree(FT)or Dynamic Fault Tree(DFT).However,many components in the AFDX network have been replaced by software,FT or DFT cannot meet complexity and dynamic dependencies of the software component system.In order to carry out the security analysis of the component software system,Kaiser et al.proposed the State/Event Fault Tree(SEFT)method by adding the concept of event and state to FT.Based on the SEFT,the quantitative and qualitative analysis of the top event failure and the minimum cut set sequence have already make some achievement,which can be helpful to improve the security of system effectively.In this paper,the qualitative analysis of the minimum cut set sequence and the quantitative analysis of the top event failure are mainly studied base on the SEFTs.And then the SEFTs based on the AFDX network data integrity are established and applied to do security analysis and verification.The main work includes the following:Firstly,due to the lack of strict formal semantics,the SEFT is transformed into a Deterministic and Stochastic Petri Net(DSPN),and then the logic gates in DSPN are determined in order to solve the minimum cut set sequences that cause the top event failure.According to the rules,the Petri net corresponding to the component is replaced by the place.Finally,The matrix method based on Petri net is improved,which makes it suitable for solving the minimum cut set and the automatic solution tool—min Cut is developed.Secondly,in order to analyze the top event failure of the SEFT,a probabilistic property analysis method based on Stochastic Hybrid Automata(SHA)is proposed.The SEFT are transformed into SHA for probability property analysis due to the method based on Interface-IMC can not analyze the probabilistic characteristics of the SEFT with deterministic delay events.Aiming at the logic gates and components in the SEFT,the semantics of conversion to SHA are given respectively.After that,the statistical model checking tool—UPPAALSMC is used to analyze the probability-property,and an example is present to be compared with the probability-property analysis method based on Interface-IMC.Finally,based on the analysis of the structure of the switch and the end system in the AFDX network and the data frame transmission process,the corresponding SEFT is established.The SEFT are applied to model the internal failure behavior of software components in AFDX network accurately.Then the model is transformed into the DSPN and the SHA to do the minimum cut set and the top event failure analysis.Through the quantitative and qualitative analysis of the data integrity of AFDX network,the structure and components can be optimized according to the analysis results.