A machine learning-based compressive spectrum sensing in 5G networks using cognitive radio networks

In recent times, evolution of communication technology and standard has grown in leaps and bounds from a conventional 1G communication technology towards the recent 5G and 6G technologies in a very short span of time. However, due to increasing scarcity of spectrum for these devices, cognitive radio networks (CRNs) have emerged to be promising solutions to allocate the required spectrum to the users in an intelligent manner. The method of compressive sensing-based cyclo-stationary feaure detection method is implemented based on a powerful CNN classifier to detect the presence or absence of PU activity. Detection probability and MSE have been improved, and the optimal detection has been reformulated for minimizing the possibility of error. Performance metrics have been compared against benchmark methods and superior performance reported. The sensing conduction and the accuracy of the proposed design are increased as 98.5%.