Evaluation of CAC Methods under non-Poisson arrival traffic in cellular networks

The mean remaining time (MRT) and guard channel with queue (GCQ) call admission control (CAC) methods are evaluated in the context of non-Poisson arrival traffic conditions. The MRT scheme uses statistical estimates of the traffic profile when admitting/rejecting new calls to meet predefined QoS levels. MRT provides network operators with a wide working interval. It is simple to compute and easy to implement, and it complies with criteria of relevance in choosing a CAC policy. The GCQ scheme has been studied extensively because of its high practical value, but always under the assumption of Poisson-distributed call arrivals. Analytical studies and field measurements show that the arrival process in wireless cellular networks is not memoryless. Therefore, the main goal of the research reported in this paper is to evaluate the performances of the two call admission control methods under realistic traffic scenarios. Our results show that handover performance is highly dependent on inter-arrival times and on the channel holding time when arrival traffic is smooth. These results have important implications for cellular network dimensioning.