Fig. 8From: Computation offloading strategy based on deep reinforcement learning for connected and autonomous vehicle in vehicular edge computingAverage offloading failure rate in every 20 episodesBack to article page