Integrated Sensing-Computation Model for Cooperative Spectrum Sensing

Cooperative spectrum sensing can effectively enhance the performance of spectrum sensing. Collaboration among secondary users plays a critical role in enhancing the sensing performance, as the degree of cooperation directly impacts the accuracy and reliability of spectrum sensing. And spectrum sensing incurs significant sensing costs, such as sensing delay, computational resources, etc. However, many existing algorithms tend to focus solely on sensing performance, while neglecting the impact of sensing cost on the overall efficiency of spectrum utilization. Therefore, our proposed method utilizes the cross-entropy learning algorithm to optimize both sensing performance and sensing cost, seeking to identify the best combination of secondary users for integrating sensing results. By taking into account for the trade-off between sensing performance and cost, our approach can effectively improve the efficiency of cooperative spectrum sensing, while maintaining high levels of accuracy and reliability.