Reliability of time-constrained multi-state network susceptible to correlated component faults

Correlation can seriously degrade reliability and capacity due to the simultaneous failure of multiple components, which lowers the probability that a system can execute its required functions with acceptable levels of confidence. The high cost of fault in time-critical systems necessitates methods...

Full description

Saved in:
Bibliographic Details
Published in:Annals of operations research Vol. 311; no. 1; pp. 239 - 254
Main Authors: Lin, Yi-Kuei, Fiondella, Lance, Chang, Ping-Chen
Format: Journal Article
Language:English
Published: New York Springer US 01.04.2022
Springer
Springer Nature B.V
Subjects:
Business and Management
Combinatorics
Component reliability
Computer networks
Constraint modelling
Design optimization
Electronic data processing
Fault location (Engineering)
Faults
Methods
Network reliability
Operations research
Operations Research/Decision Theory
Reliability (Engineering)
S.I.: Reliability Modeling with Applications Based on Big Data
System reliability
Theory of Computation
Taiwan
System reliability
Time-constrained multi-state network (TCMSN)
Correlated fault
ISSN:0254-5330, 1572-9338
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Correlation can seriously degrade reliability and capacity due to the simultaneous failure of multiple components, which lowers the probability that a system can execute its required functions with acceptable levels of confidence. The high cost of fault in time-critical systems necessitates methods to explicitly consider the influence of correlation on reliability. This paper constructs a network-structured model, namely time-constrained multi-state network (TCMSN), to investigate the capacity of a computer network. In the TCMSN, the physical lines comprising the edges of the computer network experience correlated faults. Our approach quantifies the probability that d units of data can be sent from source to sink in no more than T units of time. This probability that the computer network delivers a specified level of data before the deadline is referred to as the system reliability. Experimental results indicate that the negative influence of correlation on reliability could be significant, especially when the data amount is close to network bandwidth and the time constraint is tight. The modeling approach will subsequently promote design and optimization studies to mitigate the vulnerability of networks to correlated faults.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0254-5330
1572-9338
DOI:10.1007/s10479-019-03428-3