Published Paper Details:

The spread of wireless technology has users scrambling to find more usable radio spectrums. The search has shown inefficiencies in licensed spectrum bands and shown that they are getting harder to find. Without interfering with major users, Cognitive Radio Networks (CRNs) enable secondary users to opportunistically access unused spectrum. Spectrum sensing, which enables users to determine which frequencies are unoccupied and hence suitable for transmission, is one useful feature of CRNs. Energy detection is the most straightforward and rapid of the several spectrum sensing techniques. However, there are disadvantages to the conventional approach of energy sensing. A poor signal to noise ratio (SNR), a high rate of false alarms, and uncertainty with noise levels are some of these disadvantages under fading circumstances. These constraints have a significant impact on overall network performance and reduce the efficiency of spectrum sensing. Reliable and precise spectrum sensing utilizing adaptive threshold, noise variance, and informed detection at low SNR is the basis of our enhanced energy detection technique, which is presented in this study. The model can be improved by increasing detection probability and decreasing false alarm and missed detection probabilities. Contrasting the traditional energy detection approach with the recommended update improves communication reliability, spectrum occupancy, and spectrum quality. This study also discusses feasible enhancing application situations, issues with the proposed improvement, and future research objectives for AI and ML-based spectrum sensing.