Q-learning Based Radio Resources Allocation in Cognitive Satellite Communication
Jun Wang, Zhu Jinzhou · 2022 International Symposium on Networks, Computers and Communications (ISNCC) · 2022
The mobile satellite communication systems are surging then make the valuable wireless resources more and more rare. The common methods of resource allocation still play the primary role, but new ways are needed to find the chances of coexisting in the same frequency band. Under adequate survey, introduce a non-interrupt Q-learning based cognitive radio method and execute Monte Carlo simulation. The results show that the method can meet the interference thresholds requirement, make decisions of transmission power, information rate based on Q-learning and make the secondary users coexist with the primary-user. With priori knowledge such as learning from others or traffic information, above 60% iteration number could be reduced with less than 2% QOS or 8% bit rate expense.