Intelligent Dynamic Spectrum Allocation in MEC-Enabled Cognitive Networks: A Multiagent Reinforcement Learning Approach

Figure 5

Episode reward versus training epoch under different number of SUs.