Abstract
The effective call admission control (CAC) scheme provides preferential treatments on call requests to achieve efficient utilization of system resources. The Markov Decision Process (MDP) is one of the optimization schemes used to derive the optimal CAC policy for pre-determined goals. However, the MDP-based approaches suffer from the Curse of Dimensionality problem, where the size of state space grows exponentially with the increase of the number of call types and the number of channels. The Curse of Dimensionality problem makes the MDP-based approach difficult to be implemented in practice. In this paper, we consider a heterogeneous telecommunication system with multiple service priority classes. We propose a suboptimal MDP-based CAC scheme, designed based on a two-phase reduced dimension structure to substantially reduce the overall computational complexity from the order of O(C 12 ) to the order of O(C 4 ), where C denotes number of system capacity. Our proposed MDP-based CAC scheme is evaluated via an event driven simulator and the results are compared with the conventional complete sharing scheme and different guard channel schemes under various traffic loads. © 2012 IEEE.