Detection of Attackers in Cognitive Radio Network Using Optimized Neural Networks

Abstract

Cognitive radio network (CRN) is a growing technology targeting more resourcefully exploiting the available spectrum for opportunistic network usage. By the concept of cognitive radio, the wastage of available spectrum reduced about 30% worldwide. The key operation of CRN is spectrum sensing. The sensing results about the spectrum are directly proportional to the performance of the network. In CRN, the final result about the available spectrum is decided by combing the local sensing results. The presence or participation of attackers in the network leads to false decisions and the performance of the network will be degraded. In this work, an optimized artificial neural network (ANN) based aggressor classification algorithm is proposed. The performance of ANN improved by using the Immune plasma optimization (IPO) algorithm which is inspired by human immune system response for COVID-19 disease. Results indicate that the proposed IP optimized ANN produces better results in terms of attacker detection accuracy, energy, packet delivery ratio and delay of the network. The results show that the proposed method has 32% accuracy rate improvement’s, 16% energy savings, 40% packet delivery ratio improvements and 30% overall delay reductions than the existing methods.