Stochastic decomposition on multi-server crash in sequential revamp scheme

Abstract

Network optimization techniques have found a prime application to design the framework and operational analysis for large scale network. In the stochastic system of communication and computer network the server crash have massively more attention, for the reason that noticeably a negative impact on the performance and functionality of computer networks such as processor failure, a service disruption, job priority and some peripheral riot factor. This paper deals with multi queue network of N server may occur randomly with many types of crashes. Multiple paths may exists between the source-destination nodes that direct that traffic load variations, overhead or response time. Each types of crash require repairs a finite number of stages before the service is smarten up. Abortive servers are repair in the sequential order follows the previous stage. Moreover the traffic congestion has become a critical problem which deteriorate the Quality of Service for network users. Stochastic decomposition has employed to obtain approximations for the queue length distributions. The usefulness of a particular stochastic model depends on both its computational advantages and on the extent to which it can be adjust to describe different phenomena. Graphical representation shows that how the new method improves papers the performance measure on queueing network in multi server crash.