Fig. 6From: Computation offloading strategy based on deep reinforcement learning for connected and autonomous vehicle in vehicular edge computingAverage reward in every 20 episodesBack to article page