TY  - JOUR
T1  - Optimizing UAV Network Routing with GNNs and Transfer Learning for Low Latency and High Throughput
AU  - Lee, Seunghyeon 
AU  - Park, Changmin 
AU  - Kim, Hwangnam 
JO  - The Journal of Korean Institute of Communications and Information Sciences
PY  - 2025
DA  - 2025/1/1
DO  - 10.7840/kics.2025.50.9.1417
KW  - UAV networks
KW  - Graph Neural Network
KW  - Routing optimization
KW  - Transfer Learning
KW  - Low-latency
KW  - communication
AB  - Unmanned Aerial Vehicle (UAV) networks, while offering benefits like high mobility and line-of-sight communication, face significant challenges such as high latency and unreliable connectivity. To overcome these issues, this paper introduces a Graph Neural Network (GNN)-based routing approach leveraging transfer learning to optimize path prediction with a focus on both latency and throughput. Experimental results indicate that the proposed method outperforms Dijkstra-based routing in terms of inference speed and accuracy, especially in large-scale networks, highlighting its potential as an effective low-latency, high-throughput solution for UAV networks.