Optimization of Stochastic Performance Analysis on Multi Class Queueing Network

Queueing network models have used lucratively to analyze the validation of multi resource system such as Computer systems and Communication Networks. This paper covenantal with the Stochastic analysis of queueing network with several classes of customers. Stochastic analysis describes the nature of the queueing networks and their applications, calculating the system performance. The measurable parameters such as utilizations, mean queue length, response times are defined and the computational rapport is observed from their output generated by the simulation. Operational variables portrays the behavior of queueing network, for which in the case of large-scale queueing network, the solution may be written as a product of terms, where each term corresponding to the solution of an individual station of the network. The three basic operational assumptions such as flow balanced, routing homogeneity and onestep behavior are used to construct the analysis of network state ‘n’ as the time-proportion of queueing network P(n) . Finally, the new concept of decomposition principle has used to replace the non-homogeneous sub-system with equivalent devices. This property has proven useful to reduce the computation performance measures of the queueing networks with jobs reveal vastly unstable service load for the various service stations in the network. The model represents a new approach for analyzing transient congestion phenomena in such networks. Graphical representation shows that how the new method improves the performance measure over the simulation period in multi class queueing network.