Queue Control Model in a Clustered Computer Network using M/M/m Approach

This paper presents an application of queueing model for controlling queues in clustered computer networks by incorporating multiple servers in the system using multi-server queueing algorithm for the simulation. It has been shown through theory and experiment that as arrival rate increases in clustered computer networks, single server becomes inefficient and prone to instability, irrespective of the queueing system involved leading to unnecessary delay for service. To address this problem, we deployed an M/M/m queueing model using queueing theory and Object Oriented System Analysis and Design Methodology to incorporate multiple servers to serve customers such that as arrival rate increases, the servers remain efficient, fair and stable which leads to a reduction in waiting time. The Model is simulated using Visual C# (C Sharp) and Microsoft.NET Framework programming tool. Though the application is modeled within the context of FUTO campus network but it is a plausible approach to estimating a reduction in waiting times, traffic intensity (queue length), increasing server utilization, and can be adapted to any other networks.