Optimisation of throughput in cognitive radio networks: an analysis at the data link layer

In this study, the authors investigate the optimal sensing settings that can maximise the throughput of the cognitive radio (CR) users under the constraint that the signals from the primary users (PUs) are sufficiently protected. They focus on the throughput at the data link layer where end-to-end delivery of data packets is considered. Consequently, the throughput is defined as the rate at which correct data are received. They reconsider the tradeoff between the sensing settings and the throughput. They also separately perform optimisation on the throughput when cooperative spectrum sensing (CSS) is/is not used. In order to evaluate the throughput, they further derive the exact symbol-error-rate (SER) expressions in the presence of co-channel interference. After proving that the throughput functions are unimodal, they proposed two efficient and effective algorithms that can derive the optimal settings for maximising the throughput. Finally, the authors show that their numerical results and compare optimal results with those found by exhaustive searches. They conclude that their algorithms produce optimal results very close to those found by exhaustive searches.