A QoE-Driven Spectrum Decision Scheme for Multimedia Transmissions over Cognitive Radio Networks

Spectrum decision is the key function of a cognitive radio(CR) to choose the most proper spectrum band to meet secondary users'(SUs') quality of experience(QoE) requirements. It is challenging to perform spectrum decision for multimedia applications in cognitive radio networks(CRNs) due to two reasons. One is that QoE evaluation model for SU in CRNs is still an open issue, for the conventional QoE models neglecting the effects of primary users'(PUs') behavior. The other reason is the dynamic characteristics of available channel and competitions among SUs cause difficulty to satisfy heterogeneous SUs' QoE requirements. Therefore, firstly we develop a satisfaction probability-based QoE evaluation model for multimedia CRN taking specific metrics handoff delay and handoff frequency into account. Then, based on this model, we proposed a spectrum decision scheme to maximize the SUs' expectation of mean opinion score(MOS). The scheme adaptively achieve spectrum decision according to changing channel condition and SU traffic load. Simulation results demonstrate that the proposed QoE-driven spectrum decision scheme improves SUs' QoE satisfaction.