Fault-tolerant computing concepts for aerospace applications- a survey

Abstract

A fault-tolerant computing system performs its intended functions irrespective of the occurrence of certain failures. As the system becomes more and more reliable with built-in fault-tolerance, its analysis, validation and comparison with another system become formidable tasks. Such a study involves issues such as fault classification, figures of merit, fault-tolerant architectures, coverage estimation and automated methods of reliability evaluation. This paper provides a comprehensive survey of all these topics. The new concept of performability, which combines both the performance and the reliability of a system, and the configuration optimisation of a gracefully degradable computing system are also discussed.