ESTIMATION METHODS IN EVALUATION AND DESIGN OF RELIABLE NETWORKS:APPLICATIONS AND SCOPE

摘要

Reliable networks are the ones that take the failure prone nature of network elements into account. They are used to model a wide variety of real life problems. There are two interesting research problems in the area of reliable networks, the problem of quanitfying the network performance in terms of a performance metric, and given a performance metric, the problem of designing a reliable network that meets certain pre-specified design criteria. Both of these problems are NP-hard, therefore exact solution methods are of limited use. Real life instances of these problems are solved using Monte Carlo estimation methods. In this paper, we review the various Monte Carlo methods used for estimating the performance of a reliable network and point out directions for future research.