Agentic Graph Neural Networks for Wireless Communications and Networking Toward Edge General Intelligence: A Survey

Abstract

The rapid advancement of communication technologies has driven the evolution of communication networks toward both high-dimensional resource utilization and multifunctional integration. This evolving complexity poses significant challenges in designing communication networks to satisfy the growing quality-of-service and time sensitivity of mobile applications in dynamic environments. Graph neural networks (GNNs) have emerged as fundamental deep learning (DL) models for complex communication networks. Most existing GNNs are task-specific, whereas end-to-end communication performance hinges on multi-step inference. To address this gap, this article proposes to leverage agentic artificial intelligence (AI) to orchestrate and integrate diverse GNNs, thereby forming a novel framework termed agentic GNNs. This framework enables application-aware implementations, facilitating the advancement of edge general intelligence. Regarding the core roles of GNNs in the framework, we comprehensively review recent advances in GNN-based applications for wireless communications and networking, aiming to fully understand the comprehensive capabilities of GNNs. Specifically, we focus on the alignment between graph representations and network topologies, as well as between neural architectures and communication tasks. We first provide an overview of GNNs based on prominent neural architectures, followed by the concept of agentic GNNs. Then, we summarize and compare GNN applications for conventional systems and emerging technologies, including physical, MAC, and network layer designs, integrated sensing and communication (ISAC), reconfigurable intelligent surface (RIS) and cell-free network architecture. We further propose a large language model (LLM) framework as an intelligent question-answering agent, leveraging this survey as a local knowledge base to enable GNN-related responses tailored to wireless communication research. Moreover, we present several experimental results to quantify the effectiveness of GNNs across various scenarios. Finally, we highlight the critical challenges, open issues, and future research directions for GNN-empowered wireless communication designs.