Reliability for a Stochastic Flow Computer Network subject to Time Constraint and Correlated Failures  
Author Ping-Chen Chang

 
Co-Author(s) Yi-Kuei Lin; Ding-Hsiang Huang

 
Abstract This paper constructs a time-based stochastic flow network (TSFN) model to evaluate the performance of a computer system. In such a TSFN, the failure of the physical lines comprising the edges of the computer network can experience correlated failures. In particular, the data transmitted by the network can be sent through two minimal paths simultaneously. A quantitative approach is proposed to evaluate the probability that a given amount of data can be sent from source to sink within a given time. The probability is defined as the system reliability to indicate the performance of the computer system. Experimental result shows that the impact of correlation on reliability could be significant under different constraints. The modeling approach can be further applied to optimization problem.

 
Keywords Correlated failure, time-based stochastic flow network (TSFN), system reliability, two minimal paths
   
    Article #:  RQD25-24
 
Proceedings of 25th ISSAT International Conference on Reliability & Quality in Design
August 1-3, 2019 - Las Vegas, NV, U.S.A.