Optimal integrated call admission control and dynamic pricing with handoffs and price-affected arrivals

Abstract

Call admission control and dynamic pricing have been proposed as separate measures to reduce congestion in a network. In this paper, we integrate the call admission control and dynamic pricing problems, formulating them as a Markov Decision Problem (MDP), operating in a multiservice, resource-sharing cellular system. Our model incorporates price-affected behaviour of network users, considering effects on arrivals, retrials and substitutions among services and time. The network exercises admission control, to reject or accept a new connection requests, and price control to set a state-dependent price per bandwidth time with the objective to maximise the long-term revenue. The results show that the proposed method improves revenue and reduces congestion compared to other conventional sub-optimal policies. © 2005 IEEE.