Vacation policies in an M/G/1 type queueing system with finite capacity (Q1109436): Difference between revisions

An M/G/1 queueing system is considered. The system has a finite capacity L, so the customer finding L customers present in the system may not enter it. Each time the number of customers in the system drops to \(\nu\geq 0\), the server stops serving the queue and remains inactive for a certain period of time. The system starts at time \(t=0\) with a departure and \(\nu\) customers left in the system. Then an inactive phase begins for the server who becomes active again at time \(\tau_ 1\). This active phase ends when the number of customers drops again to \(\nu\). Thus the server is alternately in active and inactive states. Let \(v_{n\nu}\) denote the duration of the n-th inactive phase. Let \(\pi\) \(L_{v\nu}(j)\) and \(p\) \(L_{v\nu}(j)\), \(j=\nu\),..., L-1, denote the stationary distribution of the number of customers in the system immediately after the n-th departure and at the arbitrary moment t, respectively. Let \(\pi\) L(j) and p L(j) denote the respective distributions of the classical model M/G/1/L. Formulae relating \(\pi\) \(L_{v\nu}\) and \(\pi\) L and also \(p\) \(L_{v\nu}\) and p L are given. Some particular cases are also investigated.