Title :
Reliability prediction for a class of highly reliable digital systems
Author_Institution :
NASA Langley Res. Center, Hampton, VA, USA
Abstract :
Three theorems which show that the reliability of a popular class of systems can be computed using small and simple models are presented. This class consists of systems that are assemblages of subsystems where each subsystem is a majority-voting threeplex plus spares or majority-voting fourplex plus spares. The theorems are error bounds for model reduction and simplification. The error bounds are given in terms of readily available system parameters. The three theorems have been applied to a system that has been used as an example that generates extremely large reliability models; the system considered is one version of AIPS (Advanced Information Processing System) for IAPSA (Integrated Airframe Propulsion System Architecture)
Keywords :
aerospace computer control; aerospace propulsion; failure analysis; reliability; AIPS; Advanced Information Processing System; IAPSA; Integrated Airframe Propulsion System Architecture; aerospace propulsion; digital systems; error bounds; majority voting; models; redundancy; reliability; spares; Aerospace control; Computer architecture; Digital systems; Failure analysis; NASA; Predictive models; Reduced order systems; Redundancy; Reliability theory; System recovery;
Conference_Titel :
Reliability and Maintainability Symposium, 1992. Proceedings., Annual
Conference_Location :
Las Vegas, NV
Print_ISBN :
0-7803-0521-3
DOI :
10.1109/ARMS.1992.187794