Detection and Performance Improvisation Strategy-Enhancing Primary User Signal Integrity in Noisy Environment

Cognitive radio (CR) is a smart radio network that relies heavily on spectrum sensing. The prime user (licensed user) is the only one who has permission to access the spectrum. The primary user (PU) signal is often constrained by additive white Gaussian noise (AWGN). To detect primary signals available at low SNR (<−6db), the deep unfolding strategy allows unlicensed secondary users to exploit the prime user’s unused spectrum. The availability of PU signal using the set of DeepSig datasets can be detected by the energy detection hypothesis and a better-performing classifier is analysed using different classifiers. In the deep unfolded approach, the probability of signal available up to SNR=−10 is evaluated, and the decision tree classification algorithm with an accuracy of 84.9% is chosen as the best classifier for this DeepSig datatype.