Fig. 7From: Reinforcement learning-based hybrid spectrum resource allocation scheme for the high load of URLLC servicesThe delay curves at various loads. The delay performance curves of the RL-based resource allocation scheme with loads of 0.86, 1.03, and 1.28. The x-axis is the number of iterative trainings, and the y-axis is the average delay. The total delay of URLLC data at various loads is decreased by the number of training increases and converges to a certain value after limited trainingBack to article page