A GNN-Based Supervised Learning Framework for Resource Allocation in Wireless IoT Networks

The Internet of Things (IoT) allows physical devices to be connected over the wireless networks. Although device-to-device (D2D) communication has emerged as a promising technology for IoT, conventional solutions D2D resource allocation are usually computationally complex and time consuming. high complexity poses significant challenge practical implementation IoT A graph neural network (GNN)-based framework is proposed address this in supervised manner. Specifically, modeled directed graph, where desirable links nodes harmful interference edges. effectiveness verified via two case studies, namely link scheduling networks joint channel power underlaid cellular Simulation results demonstrate that outperforms benchmark schemes terms average sum rate sample efficiency. In addition, GNN approach shows potential generalizability different system settings robustness corrupted input features. It also accelerates optimization by reducing execution only few milliseconds.