Voice-Service Capacity Analysis for Cognitive Radio Networks

In this paper, quality-of-service (QoS) provisioning for voice service in cognitive radio networks is considered. As voice traffic is sensitive to delay, the presence of primary users and the requirement that secondary users should not interfere with them pose many challenges for QoS support in cognitive radio networks. Two cognitive medium-access control (MAC) schemes are proposed in this paper for secondary voice users to access the available channel. One is the contention-based scheme, and the other is the contention-free scheme. An analytical model is developed to obtain the voice-service capacity (i.e., the maximum number of secondary voice users that can be supported with QoS guarantee) of the two proposed schemes, taking into account the impact of the primary users' activity. Both independent and correlated channel busy/idle state models for primary activity are considered. The analytical model is validated by simulations. The analytical results will be useful to support voice service in cognitive radio networks.