Solving the Deadlock Problem with Deep Reinforcement Learning Using Information from Multiple Vehicles | IEEE Conference Publication | IEEE Xplore