Research on Multi-user Dynamic Spectrum Allocation Strategy Using Reinforcement Learning in Unknown Environments
Abstract— With the advent of 5g era, the demand of spectrum expands, which leads to the shortage of spectrum resources. Because of the low utilization of spectrum resources, it is very important to find efficient spectrum allocation strategy for wireless communication. Cognitive Radio technology will become the key to solving this problem . This paper proposes a multi-user system model that conforms to the actual 5G communication situation, and uses the reinforcement learning Deep-Q-Network (DQN) algorithm to study its dynamic spectrum allocation problem. The simulation results show that the algorithm under this model can converge quickly and improve the efficiency of spectrum resource utilization.
Index Terms— Cognitive Radio, multi-user, 5G, DQN, dynamic spectrum allocation
Shilong Cao, Fei Lin, Xianzhi Jin
Qilu University of Technology (Shandong Academy of Sciences), CHINA
Cite: Shilong Cao, Fei Lin, Xianzhi Jin, "Research on Multi-user Dynamic Spectrum Allocation Strategy Using Reinforcement Learning in Unknown Environments, " Proceedings of 2021 the 11th International Workshop on Computer Science and Engineering (WCSE 2021), pp. 22-30, February 25-27, 2021.