Abstract: Communication networks are difficult to model and predict because they have become very sophisticated and dynamic. We develop a reinforcement learning routing algorithm (RLRouting) to solve ...