Performance analysis of cognitive radio networks using channel-prediction-probabilities and improved frame structure

This paper focuses on potential issues related to the random selection of a sensing channel that occurs after the prediction phase in a Cognitive Radio Network (CRN). A novel approach (Approach-1) for improved selection is proposed, which relies on the probabilities of channels by which they are predicted idle. Further, closed-form expressions are derived for the throughput of Cognitive Users (CUs) in the conventional and proposed approaches. In addition to this, a fundamental approach for computing the prediction probabilities is also proposed. Moreover, a new challenging issue named “sense and stuck” was observed in the conventional approach. The proposed approach is validated by comparing the results achieved with the results of the conventional approach. However, to achieve the prediction probabilities, the pre-channel-state-information is a prerequisite, but it may be unavailable for particular scenarios; therefore, a modified selection method is introduced to avoid the sense and stuck problem. An algorithm to evaluate the throughput using the random, improved, and modified selection methods is presented with its space and time complexities. Furthermore, for additional improvement in the throughput of the CU, a new frame structure is introduced, in which the spectrum prediction and sensing periods are exploited for simultaneous transmission of data via the underlay spectrum access technique (Approach-2). The simulated results of Approach-2 are compared with our pre-obtained results of Approach-1, which confirm significant improvement in the throughput. Keywords: Cognitive radio, Interweave, Prediction, Spectrum accessing, Throughput, Underla