This study details how a cognitive radio network and a wavelet transform can be used to build a compressive sensing technique for locating gaps in the spectrum. IoT deployments have limitations in areas such as resilience in hostile environments, bandwidth allocation and utilization, ease of use, and RF spectrum pricing. Effective use of the given RF spectrum is one of the greatest ways to integrate IoT in 5G, as this spectrum is typically underutilized due to consumption by the licensed users known as Primary Users (PUs). As a result, the Spectrum Decision by CR's unlicensed users is important for CR-based IoT in networks supporting 5G and beyond. Bayesian Compressive Sensing is utilized here to deal with the process's inherent complexity and uncertainty. Since the proposed Bayesian RVM-based compressive sensing technique, Bayesian Compressive Sensing requires less information about the measurement noise. Remarkably high accuracy and speed are maintained despite the fact that this technique requires fewer measurements to recover wideband signals. Even in transmissions from unlicensed users, who are subject to little regulation, the wavelet transform is used in this work to detect the primary user (PU). The method's value lies in the fact that it permits simultaneous evaluation of all viable hypotheses within the global optimization framework. Research is done using simulations to test how well the proposed method works in a cognitive radio setting. Recovery error, recovery time, and covariance are compared to the standard Bayesian approach to show the superiority of the proposed method.

Internet of Things, Cognitive radio, Bayesian compressive sensing, Wavelet Transform, Relevance Vector Machine, fifth generation (5G).