Performance Analysis of M/G/1 Retrial Queue with Finite Source Population Using Markov Regenerative Stochastic Petri Nets

This paper aims to present an approach for modeling and analyzing an M/G/1//2 retrial queue, using the MRSPN ( Markov Regenerative Stochastic Petri Nets ) tool. The consideration of the retrials and finite source population introduce analytical di culties. The expressive power of the MRSPN formalism provides us with a detailed modeling of retrial systems. In addition to this modeling, this formalism gives us a qualitative and a quantitative analysis which allow us to obtain the steady state performance indices. Indeed, some illustrative numerical results will be given by using the software package Time Net.