Traffic signal control by distributed Reinforcement Learning with min-sum communication | IEEE Conference Publication | IEEE Xplore