Modeling and availability analysis of computer networks using fuzzy reliability approach

The traditional theory of reliability is based on the Bernoulli trials, i.e., either success or failure. However, itseems unrealistic for enormous complex systems like computer networks. To overcome this issue, in the present study, an effort has been made regarding the development of a mathematical model for a computer network system and analysis of fuzzy availability of the same. The concepts of constant failure, constant repair and coverage factor have been used for the development of the model. The impact of coverage factor, repair rates and failure rates of components has been analyzed on the fuzzy availability of the system. Markov birth-death process has been used for the development of Chapman-Kolmogorov differential equations. Theleading differential equations have been simplified by Runge-Kutta method of order four using MATLAB (Ode 45 function).