Effect of Shadow Cluster on Predictive User Mobility Behavior Scheme in Cellular Networks

In this paper, we propose a scheme for improved performance of resource reservation and call admission control for cellular networks called Predictive User Mobility Behavior (PUMB) scheme. This algorithm is proposed in which bandwidth is allocated more efficiently to neighboring cells by key mobility parameters in order to provide Quality of Service (QoS) guarantees for transferring traffic. The cell visiting probability (CVP) is used to form the shadow cluster cell, where a mobile unit is likely to visit, by comparing with shadow cluster probability threshold. We are interested in the effect of the shadow cluster probability threshold in PUMB because when shadow cluster probability threshold increases, the MU´s shadow cluster size decreases so that the suitable shadow cluster probability threshold can define the suitable shadow cluster size for increasing network performance. The performance of the PUMB is demonstrated by simulation results in terms of new call blocking probability and handoff call dropping probability when new call arrival rate and the shadow cluster threshold are varied. Our results show that when shadow cluster probability threshold increases, CBP of MSODB and PUMB decrease while CDP of both schemes increases in all arrival rates. However, PUMB is better than MSODB under different traffic conditions.