An improved ensemble machine learning classifier for efficient spectrum sensing in cognitive radio networks

Summary Cognitive radio network (CRN) is one form of wireless communication for solving the spectrum underutilization problem. It is mainly used for sensing and learning the electromagnetic environment and changes according to the environment. CRNs mainly depend on the cooperation functionality for making the network work efficiently. The main technology in cognitive radio is spectrum sensing. The cooperative spectrum sensing (CSS) scheme is used for estimating the high transmit power in the cognitive radio networks, and the result is that it satisfies the interference constraints. It develops the communication overhead for the local observation dissemination between the secondary users. It is used for speed and accurate detection techniques of the user, and it also identifies the spectrum holes without any interference to others while sharing with other users. It is based on the ensemble support vector machine (SVM) with CSS for producing high performance in CRNs with the use of artificial intelligence (AI) techniques. Primary users are licensed users, and secondary users are unlicensed users. Secondary users are searching and sensing the availability of the primary users' channel. The following features are considered for classification: PU transmission probability, bandwidth, RSSI, distance, transmission power, and sensing time along with a class label denoting the presence and absence of PU.