Fig. 6From: A deep reinforcement approach for computation offloading in MEC dynamic networksAverage delay versus task requirementBack to article page