AI-based D2D Network Managment

we evaluate resource management in device-to-device (D2D) networks that can coexist with mobile cellular user equipments (CUEs). Moreover, a new smart model for joint scheduling and resource management for D2D networks is presented. In this regard, aiming to guarantee the freshness of information, which is equivalent to minimizing the average age of information (AoI), and to utilize energy harvesting (EH) technology to address energy requirement for D2D network, we formulate an online optimization problem with quality of service (QoS) of CUEs and D2Ds, available power, freshness of information, and environmental requirement constraints. Since the proposed optimization problem belongs to mix-integer nonlinear problems, and it is online, we adopt a deep reinforcement learning (DRL) based solution to solve it efficiently.