Fig. 4From: Deep reinforcement learning-based adaptive modulation for OFDM underwater acoustic communication systemThrough main network output, the Q values of all actions corresponding to a CSI stateBack to article page