Fig. 3From: Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approachLocal observation and decentralized decision of user agent at slot tBack to article page