A Framework for Secure Cooperative Spectrum Sensing based with Blockchain and Deep Learning model in Cognitive Radio

Today we live in era where not only humans interact but machines interact too. Internet of Things has disrupted the communication with an enormous growth in number of connected devices worldwide. This resulted in big challenges to meet the spectrum requirement of these devices such as seamless connectivity, scalability and accessibility. Cognitive Radio (CR) is designed to meet the requirement since it uses spectrum holes in the licensed bands. Security issues put at risk spectrum sensing, a crucial part of the Cognitive Radio Network (CRN).A malicious user (MU) reduces the accuracy of spectrum sensing, particularly in the situation of cooperative spectrum sensing where MU transmits fabricated data to the fusion centre. The performance of cognitive radios may suffer from the presence of such MU in the system that create erroneous sensing data. As a result, this paper proposes a Blockchain-based method for MU detection in networks. This strategy makes it simple to distinguish between a trustworthy user and a MU using cryptographic keys. The effectiveness of the suggested technique is examined using python tool. The proposed method detects Malicious user with 100 % efficiency in very less sensing time of 0.6ms. The results were also compared with adaptive threshold, FOF and TTA algorithms.