A QoS-aware cross-layer scheduling scheme for multiuser mixed-traffic cognitive radio networks

The emerging technology of cognitive radio (CR) allows secondary users (SUs) to gain access to radio spectrum which is licensed to but temporarily unoccupied by primary users (PUs). Apart from spectrum sensing, dynamic spectrum access (DSA) is an important processing task for CR networks (CRNs). Its aim is to allocate spectrum for SUs dynamically without causing any harm to PUs. CRNs may support different traffic types that may have different quality of service (QoS) requirements. In order to meet these QoS requirements, channel states, queue states and traffic types should be taken into account in designing DSA schemes. However, traffic types handled by SUs are treated equally in most existing DSA schemes. We propose a cross-layer scheduling scheme by incorporating DSA, packet scheduling and a channel sharing policy to achieve efficient QoS support for SUs with different traffic types. QoS performance of the proposed scheduling scheme is evaluated in terms of average throughput, average packet delay and packet dropping probability for SUs with different traffic types. Simulation results show that the proposed scheduling scheme outperforms the conventional scheduling scheme by providing more efficient QoS support for SUs and higher throughput fairness among SUs as the PU's activity factor increases.