Deep Reinforcement Learning-Based Power Allocation for Minimizing Age of Information and Energy Consumption in Multi-Input Multi-Output and Non-Orthogonal Multiple Access Internet of Things Systems