Maximizing Cognitive Radio Networks Throughput Using Limited Historical Behavior of Primary Users

Cognitive radios (CRs) mainly aim to reuse the spectrum holes in order to efficiently utilize the available scarce radio spectrum. However, current CRs techniques have a throughput limitation problem which ultimately limits telecommunication applications horizons nowadays. Moreover, achieving high throughput will overcome the bottleneck of CRs application limitations to the reporting and browsing applications only. To tackle this emerging throughput limitation issue in the CRs, this paper proposes the online greedy throughput maximization (OGTM) algorithm which overcomes the throughput limitations. OGTM allows the sensing cycle frame to have a variable length according to the assumed decision validity interval. Then, OGTM varies the decision validity interval of secondary users (SUs) based on the primary users (PUs) historical behavior. As a proof of concept, we developed a simulator in order to evaluate the performance of the proposed OGTM technique. The simulation results show that SUs benefit from the limited PU historical behavior learning, which resultantly increases the throughput up to 95% and at the same time decreases the miss detection probability by 50%.