Imperfect spectrum sensing for partial spectrum-shared licensed networks

This study considers a scenario where two licensed networks share spectrum partially and opportunistically based on sensing of cognitive radios. The authors´ main contribution is a discrete-time Markov chain analysis characterising the impact of imperfect sensing and sensing periodicity on the network performance. A cross-layer design framework is also developed, where the parameters in spectrum sensing and spectrum sharing are jointly optimised. Results show that the proposed policy achieves a higher throughput than conventional strategies.