This is the complimentary code for the WIDECOMM 2023 Paper Submission: Clustering and Communication Efficiency with GNNs in the Internet of Vehicles This code contains the K-Means Clustering model with three GNN approaches: GAT, GCN, and GraphSAGE. It compares against a baseline approach with just K-Means. To use a specific GNN, you must modify the encoder to reflect these parameters.
Enjoy!