Colin Perkins and Andy Tyrrell
Proceedings of the 2nd IMA Conference on the Mathematics of Dependable Systems,
York, UK,
September 1995.
We present a new reliability model for hard real-time systems. This
is an extended Markov model, derived from an analysis of the generic
properties of hard real-time systems subject to a simple random-fault
model. Our model permits analysis of the run-time behaviour of a
system, in order to derive the probability profiles of the system's
completion/failure times. The model is applied to the analysis of a
simple sequential recovery block system, and illustrative examples
based on this system are provided. The paper concludes with a
discussion of the application of such accurate completion profile
information to the design of embedded software systems.
Download: MDS-95.pdf