KCS 2 TM: A key sharing approach and Deep Learning Model for Primary User Emulsion Attack detection in Cognitive Radio Networks

Abstract

The demand for wireless transmission and existence of the cognitive radio networks (CRN) insist the need for detecting the Primary User Emulsion Attack (PUEA) that authenticates the network as the licensed user for accessing the spectrum in CRN. Such condition creates disruptions in the communications. Hence, in this research, PUEA detection in the CRN is proposed using the deep learning and key sharing approach (KCS²TM model). The key sharing approach enables sharing the secret key between Primary User (PU) and Secondary User (SU) that supports the generation of the communication data, which easily assists in detecting the PUEA for which the ChimSp optimizer enabled Bidirectional Long Short Term Memory (CS²TM) model is proposed. The ChimSp optimizer is designed theoretically and mathematically to train the classifier and ensure the improved detection performances. The model provides increased accuracy, highly reliable, reduce computational cost and time are the supremacy of the model. The performance of the model is analyzed with the existing methods, which shows better accuracy sensitivity and specificity of 85.01%, 87.61% and 82.42% compared to other state of art methods.