It is unrealistic to assume that a computer system will never fail. Yet, there is no fault-free system. Nevertheless, even in the presence of possible faults, the system should continue operating as normal as possible and it is not expected to crash. That assertion is especially true for hard real-time systems on which human lives may depend on. Considering this scenario, this paper introduces a computational and conceptual model addressing the fault resilience improvement in a class of hard real-time system named avionics systems. Essentially, the proposed computational and conceptual model consists of a fundamental schedulability analysis and three other different algorithms. This paper shows the improvement of fault resilience, on up to 80%, by composing and applying space and time redundancy (i.e., multiprocessors and scheduling) with task migration among processors in hard real-time systems' design.
Avionics hard real-time systems' concerning fault tolerance
01.10.2012
572927 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Software fault tolerance for real-time avionics systems
NTRS | 1983
|Fault tolerance techniques for avionics wavefront array processors
Tema Archiv | 1990
|