This thesis deals with the so-called T-policy and Q-policy for a queueing system with poisson arrival process and the batch-type service mechanism. The service in Q-policy is provided when the number of customers in queue is at least greater than Q, while the service in T-policy is provided at every fixed time T. Therefore, T and Q are the dicision parameter, the analyses of which were made from both the individual and the combined Queueing control models. Under the assumptions that batch-service times are independently and identically distributed random variables and the service capacity in each batch service is infinite.
Various service mechanisms were analyzed; namely, the cases of which arrival processes can't be reviewed continuously, the possible server utilization for other system services during idle time, and service can be provided after or till fixed time T.
For each of those service mechanisms, the associated service system cost function was derived under the assumption that each batch-service charge is constant K and waiting cost per customer per unit time is also constant h.
Then, the optimal $T^*$ and $Q^*$ were computed, which minimizes the long-run expected average cost.
From the result of analyses, it was found that in both the cases of serving at every T and of making the idle time utilization when no customer is arrival, Q-policy is more preferable then T-policy. Moreover, the simultaneous consideration of both T-and Q-policy resulted in less expensive system cost than each separate policy, under the condition that the review from the idle time utilization is greater than zero.