The IEEE Transactions on Network Science and Engineering is committed to timely publishing of peer-reviewed technical articles that deal with the theory and applications of network science and the interconnections among the elements in a system that form a network. In particular, the IEEE Transactions on Network Science and Engineering publishes articles on understanding, prediction, and control of structures and behaviors of networks at the fundamental level.  The types of networks covered include physical or engineered networks, information networks, biological networks, semantic networks, economic networks, social networks, and ecological networks. Aimed at discovering common principles that govern network structures, network functionalities and behaviors of networks, the journal seeks articles on understanding, prediction, and control of structures and behaviors of networks. Another trans-disciplinary focus of the IEEE Transactions on Network Science and Engineering is the interactions between and co-evolution of different genres of networks. The core topics covered include: Network Sampling and Measurement; Learning of Network Topology; Modeling and Estimation of Network Dynamics; Network Inference; Models of Complex Networks; Modeling of Network Evolution; Network Design;  Consensus, Synchronization and Control of Complex Networks;  Interactions between and Co-evolution of Different Genres of Networks; Community Formation and Detection; Complex Network Robustness and Vulnerability; Network Interdependency and Cascading Failures; Searching in Complex Networks; Information Diffusion and Propagation;  Percolation and Diffusion on Networks;  Epidemiology in Complex Systems. 

Special Issue on Advanced Application of Graph Representation Learning in Communication Networks
截稿日期: 2026-01-01

The rapid advancement of emerging communication technologies, such as flexible antennas and integrated sensing and communication, has driven the evolution of communication networks toward both high-dimensional resource utilization and multifunctional integration. This evolution poses challenges in efficiently designing communication networks to satisfy the quality-of-service and time-sensitive requirements of mobile applications in dynamic environments. In view of the notable successes of deep learning (DL) in numerous application fields, applying DL to enhance network intelligence has become an inevitable trend. DL can be integrated into communication networks across almost all layers to enhance resource utilization efficiency and network security. Furthermore, DL serves as a key enabler that gives rise to innovative wireless applications, such as semantic communications.

The task-specific characteristic of DL necessitates the customization of learning strategies and neural networks for communication networks. Communication networks comprise core components that naturally form a graph structure, such as wireless topologies and routing patterns. Therefore, graph representation learning (GRL) has emerged as a powerful DL tool to re-design communication networks. Particularly, graph neural networks (GNNs) have been increasingly recognized as one fundamental model for graph-structured communication networks. GNNs not only augment the extraction of features over the network topology but also facilitate scalable and distributed computation. GRL/GNNs can be integrated with existing wireless DL approaches as a network feature extractor. For instance, the superiority of GNNs over other neural networks has been widely demonstrated in addressing wireless optimization problems following the “learning-to-optimize” paradigm to enable real-time and scalable inference of (near-)optimal resource allocation. GRL/GNNs shift traditional iterative optimization methods to data-driven and model-based strategies, introducing unprecedented flexibility in addressing complex problems. Exploring the advanced application of GRL/GNNs in upcoming 6G can support a variety of fundamental and practical communication designs and optimization schemes, including but not limited to

    Real-time Optimization: GRL/GNNs can automatically learn to approximate the optimal mapping from network topologies to targeted resource allocation strategies, routing results, etc., thereby enabling real-time computation.
    Scalable and Distributed Implementation: GRL/GNNs are scalable to elements in communication networks, enabling signal processing in ultra-dense and dynamic wireless environments. The scalability also facilitates distributed implementation by integrating the signaling exchange of wireless networks with the message-passing mechanism of GNNs.
    Unified Solution Framework: Learning on a graph modeling a specific network topology can extract critical topological features. The features can be harnessed to tackle diverse problems via a unified framework. Fine-tuning pre-trained GNNs via transfer learning can further enhance the solving performance.
    Complex Problem Solver: Data-driven DL offers a transformative paradigm to tackle complex communication problems through GRL/GNNs. This novel paradigm for constructing problem solvers balances the trade-off between performance and efficiency of heuristic algorithms.
    Wireless Intelligence Module: By casting GNNs as a network topology-aware feature extraction module, they can be seamlessly integrated into other wireless DL architectures (such as wireless generative artificial intelligence (GAI) and wireless large language models (LLMs)) to jointly support wireless intelligence.

While GRL/GNNs can realize intelligent signal processing, there remains a pressing need for advanced model architectures and effective mechanisms to enhance the learning performance and tackle the challenges such as guaranteeing feasible solutions, capturing heterogeneous relationships for diverse wireless services, achieving robust information transmission, enabling efficient deployment strategies, etc. Besides, the lack of standardized datasets, task definitions, and unified evaluation metrics poses challenges to the reproducibility of experimental procedures in DL-enabled signal processing. Therefore, an in-depth exploration of the connection between the DL approach and the communication tasks is urgently required for the application of GRL/GNNs in communication networks.

The integration of DL and communication networks demonstrates a cutting-edge field with revolutionary potential in future mobile communication systems. This special issue is dedicated to exploring the breakthroughs in the advanced applications of GRL in communication networks from both the theoretical and practical perspectives, in order to provide a more comprehensive understanding of the philosophy of DL-enabled communication networks.

Topics of interest include, but are not limited to:

    Theoretical fundamentals of the application of GRL/GNNs in communication networks
    Performance analysis of GRL/GNN-enabled communication networks
    Distributed implementation strategies GRL/GNN-based communication schemes
    Datasets and evaluation metrics for GRL/GNN-enabled communication networks
    Integration of over-the-air computation and GNNs
    Training strategies, activation functions, and loss functions for GNN-based signal processing
    Heterogeneous GNNs for communication networks
    LLMs and GAI with GRL in communication networks
    GRL/GNNs for multifunctional wireless services, e.g., integrated sensing and communication, physical-layer security, and simultaneous wireless information and power transfer
    Novel neural architectures of GNNs for communication networks
    GRL/GNN-enabled semantic communications
    Deep reinforcement learning based GNNs in dynamic communication environments, e.g., unmanned aerial vehicle assisted communication systems and space-air-ground integrated networks
    Application of GRL/GNNs in channel estimation and signal detection
    GRL/GNN-enabled channel coding
    GRL/GNN-enabled multiple access schemes
    GRL/GNN-enabled routing algorithms and protocols
    GRL/GNN-based designs for flexible antennas, e.g., mobile antenna, fluid antenna system, pinching antenna, and reconfigurable intelligent surface
    GRL/GNN-enabled mobile edge computing and federated learning strategies
    Integration of GRL with quantum computing and quantum machine learning in communication networks
    The prototypes and experimental studies about the GRL/GNN-based communication networks

Submission Guidelines

Prospective authors are invited to submit their manuscripts electronically, adhering to the IEEE Transactions on Network Science and Engineering guidelines. Note that the page limit is the same as that of regular papers. Please submit your papers through the online system and be sure to select the special issue or special section name. Manuscripts should not be published or currently submitted for publication elsewhere. Please submit only full papers intended for review, not abstracts, to the ScholarOne portal. If requested, abstracts should be sent by e-mail directly to the Guest Editors.

Important Dates

Manuscript Submission Deadline: 1 January 2026
Initial Decision Date: 1 April 2026
Revised Manuscript Due: 1 May 2026
Final Decision Date: 1 June 2026
Final Manuscript Due: 15 June 2026
Publication Date: Second Quarter 2026