<p>The escalating proliferation of connected devices underscores the imperative for sophisticated spectrum management. Cognitive radio networks (CRNs) dynamically allocate unused spectrum portions to optimize network efficiency and meet specific application demands, enhancing overall quality of service (QoS). This study explores the complex domain of minimizing costs based on Quality of Service (QoS) in a Cognitive Radio Network (CRN). It addresses the difficulties arising from a dynamic real-world network environment characterized by fluctuating channel conditions, imperfect spectrum sensing, collision constraints with primary users, power allocation to secondary users, the preservation of QoS, and the maximization of data rates. The proposed real-time scheduling algorithm, integrating an orthogonal frequency division multiplexing based CRN operator with a water-filling power allocation mechanism, represents a significant advancement to achieve the objectives. Simulation results illustrate the algorithm’s efficacy in minimizing total costs for CRN operators while ensuring superior QoS compared to prevailing sensing-only and leasing-only policies, thereby demonstrating a noteworthy enhancement in the existing cost minimization paradigm, maintaining QoS, imperfect spectrum sensing, collision constraint, power allocation, spectral efficiency, scalability and rate maximization. Additionally, the performance of the proposed approach is compared with the existing scenario of [48].</p>