Resource Allocation in Uplink NOMA Systems: A Hybrid-Decision-Based Multi-Agent Deep Reinforcement Learning Approach | IEEE Journals & Magazine | IEEE Xplore