Issues in energy optimization of reinforcement learning based pouting algorithm applied to ad-hoc networks