Options
2011
Conference Paper
Titel
Self learning anomaly detection for embedded safety critical systems
Abstract
Especially in embedded systems like in the automotive domain, the amount of distributed functionality of safety critical software is a challenging problem. In this paper, the question is addressed how to ensure the correct behavior of distributed communicating software systems if the specification is weak, incomplete or wrong. The key point is the usage of existing test traces from communicating modules (e.g. network traces) as basis for system diagnosis. It is discussed how this can be implemented efficiently, e.g. based on a dependency model which is inferred from test cases.
Author(s)
Konferenz