Power Allocation in Dual Connectivity Networks Based on Actor-Critic Deep Reinforcement Learning | IEEE Conference Publication | IEEE Xplore