Network-Based Approach to Increase Logical Reliability of a Vehicle E/E-Architecture

Abstract

The increasing number of functions in vehicles and their intensive interdependencies leads to a more complex e/e-architecture. This complexity increases the probability of error and influences system reliability significantly. Therefore, it is necessary to detect risks at early stages. In course of this, new methods are needed to carry out a reliability-oriented transfer of logical functions to the technical architecture. This paper presents an approach that derives a set of key figures for assessing logical reliability of functional concepts by using metrics from network theory. For this purpose, a control loop is introduced. After the functional design in a model-based environment, suitable metrics are applied. Based on the reference criterion Survival Probability, structural metrics are interpreted. With reference to the structural indicators, action strategies can be traced back to system model to increase logical reliability. First results showed that the logical reliability of functional concepts were improved by conceptual adjustments and partitioning recommendations.