Performance evaluation of virtual switch with batch arrival using quasi-birth–death process

With the fast progress of market and the innovation of modern technologies, the manufacturing sector must be more flexible and upgraded. In order to respond to the dynamically changing requirements, Information and communication systems integrated first the virtualization of servers and storage and then gradually the advances in Software-Defined Networking technology. However, the benefits of these solutions are even greater when they are optimized especially by preventing the occurrence of bottlenecks and increasing utilization of available resources. In this work, we deal with the performance analysis of virtual switch which is a key component of these solutions. A markovian queuing model with finite buffer is proposed for the performance to analysis virtual switch performance. It is assumed that packets arrive at the system in batches, but served one by one. The interactions between queues are manifested by using server vacation and polling strategy. Subsequently, we analyze the packet forwarding performance, such as average packet sojourn time and blocking probability with queuing theory and solve it by using quasi-birth death process. Finally, the numerical analysis is carried out to evaluate our proposed performance model with different parameter values.