The Effect of Correlated Failure Rates on Reliability of Continuous Time 1-Out-of-2 Software
In this paper we study the effects on system reliability of the correlation over input space partitions between the rates of failure of two-channel fault-tolerant control software. We use a continuous-time semi-Markov model to describe the behavior of the system. We demonstrate via simulation that the variation of the failure rates of the channels over the partitions of the input space can affect system reliability very significantly. With a plausible range of model parameters we observed that the mean time to system failure may vary by more than an order of magnitude: positive correlation between the channel rates makes the system less reliable while negative correlation between the channel rates implies that the system is more reliable than assuming constant failure rates for the channels. Our observations seem to make a case for more detailed reliability measurements than is typically undertaken in practice.
KeywordsHazard Rate System Reliability Sojourn Time System Failure Software Reliability
Unable to display preview. Download preview PDF.
- 2.Littlewood, B., Popov, P., et al.: Design Diversity: an Update from Research on Reliability Modelling. In: Safety-Critical Systems Symposium 2001. Springer, Bristol (2001)Google Scholar
- 4.Bondavalli, A., Chiaradonna, S., et al.: Dependability Models for Iterative Software Considering Correlation among Successive Inputs. In: IEEE International Symposium on Computer Performance and Dependability (IPDS 1995), Erlangen, Germany (1995)Google Scholar
- 5.PERFORM, Möbius: Model Based Environment for Validation of System Reliability, Availability, SEcurity and Performance. User’s Manual, v. 2.0 Draft (2006)Google Scholar
- 6.Littlewood, B.: A Semi-Markov Model for Software Reliability with Failure Costs. In: MRI Symposium on Computer Software Engineering, pp. 281–300. Polytechnic Press (Available from Wiley, London), Polytechnic of New York, New York (1976)Google Scholar
- 7.David, H.A., Moeschberger, M.L.: The theory of competing risks. Griffin’s Statistical Monographs & Courses, ed. D.S.E. Prof. Alan Stuart, vol. 39, p. 103 (1978)Google Scholar
- 10.Nadarajah, S., Kotz, S.: Reliability for Some Bivariate Exponential Distributions. Mathematical Problems in Engineering, 2006, 1–14 (2006)Google Scholar
- 13.Popov, P., Littlewood, B.: The Effect of Testing on Reliability of Fault-Tolerant Software. In: Dependable Systems and Networks (DSN 2004). IEEE Computer Society Press, Florence (2004)Google Scholar
- 14.Bishop, P.G., Pullen, F.D.: PODS Revisited - A Study of Software Failure Behaviour. In: 18th International Symposium on Fault-Tolerant Computing. IEEE Computer Society Press, Tokyo (1988)Google Scholar