In this study, a controllable M / G / 1 queuing model operating under the dyadic Min ( N, T ) policy is considered. In such a queueing systems, a server is removed from the system to perform other tasks when there are no customers in the system. Then,...
In this study, a controllable M / G / 1 queuing model operating under the dyadic Min ( N, T ) policy is considered. In such a queueing systems, a server is removed from the system to perform other tasks when there are no customers in the system. Then, the removed server is reactivated immediately after the number of customers in the system reaches N for the first time or T time units elapse after removal of the server, whichever comes first. The expected busy period, which corresponds to the length of time between an activation and the next removal of the server, has been derived analytically by various ways. Since it is very difficult to apply the result to various areas and to determine the optimal values of decision variables because of its complexities. As an alternative to overcome such problems, the expected busy period is analyzed numerically in this study. For this purpose, a simple relationship model to compute the expected busy period is developed. This model could then be applied to many areas easily as well as to provide possibilities for derivation of the optimal operating policy which minimizes the total expected cost per unit time in system operations.