A Graph Foundation Model for Wireless Resource Allocation

Yucheng Sheng; Jiacheng Wang; Le Liang; Hao Ye; Shi Jin

cs.LGApr 8, 2026

A Graph Foundation Model for Wireless Resource Allocation

Authors: Yucheng Sheng, Jiacheng Wang, Le Liang, Hao Ye, Shi Jin

Organizations: National Mobile Communications Research Laboratory, Southeast University, Nanjing 210096, China · Department of Electrical and Computer Engineering, University of California, Santa Cruz, CA 95064, USA

Abstract

The aggressive densification of modern wireless networks necessitates judicious resource allocation to mitigate severe mutual interference. However, classical iterative algorithms remain computationally prohibitive for real-time applications requiring rapid responsiveness. While recent deep learning-based methods show promise, they typically function as task-specific solvers lacking the flexibility to adapt to different objectives and scenarios without expensive retraining. To address these limitations, we propose a graph foundation model for resource allocation (GFM-RA) based on a pre-training and fine-tuning paradigm to extract unified representations, thereby enabling rapid adaptation to different objectives and scenarios. Specifically, we introduce an interference-aware Transformer architecture with a bias projector that injects interference topologies into global attention mechanisms. Furthermore, we develop a hybrid self-supervised pre-training strategy that synergizes masked edge prediction with negative-free Teacher-Student contrastive learning, enabling the model to capture transferable structural representations from massive unlabeled datasets. Extensive experiments demonstrate that the proposed framework achieves state-of-the-art performance and scales effectively with increased model capacity. Crucially, leveraging its unified representations, the foundation model exhibits exceptional sample efficiency, enabling robust few-shot adaptation to diverse and unsupervised downstream objectives in out-of-distribution (OOD) scenarios. These results demonstrate the promise of pre-trained foundation models for adaptable wireless resource allocation and provide a strong foundation for future research on generalizable learning-based wireless optimization.

Figures & tables

Fig. 1: Architecture of the proposed foundation model, GFM-RA. The framework incorporates a bias projector within the graph foundation model block (right panel) to explicitly inject edge features (interference) into the attention mechanism for node representation learning.

Fig. 2: Illustration of the generative pre-training strategy based on edge prediction. The encoder reconstructs masked interference links from the partial graph topology to capture local interference patterns.

Fig. 3: The proposed contrastive pre-training framework utilizing a Teacher-Student architecture. The Student encoder predicts the target representations of the Teacher encoder, updated via EMA, to ensure robustness against topological variations.

Category	Parameter / Description	Value
Architecture	Number of Layers ( $L$ )	6
	Hidden Dimension ( $d_{\text{model}}$ )	768
	Number of Attention Heads ( $M$ )	32
Pre-training	Total Epochs	200
	Batch Size	512
	Learning Rate ( $\eta$ )	$1\times 10^{-4}$

TABLE I: List of Hyperparameters and Training Settings

Fig. 4: Performance of adapting the foundation model to three distinct utility functions under varying network densities in OOD scenarios ( $\mathcal{D}_{16}$ – $\mathcal{D}_{20}$ ).

Fig. 5: Robustness of the foundation model against the incompleteness of CSI. The aggregated performance is averaged across five OOD scenarios (

\mathcal{D}_{16}

–

\mathcal{D}_{20}

Fig. 6: Few-shot adaptation performance for sum rate maximization across OOD scenarios.

Fig. 7: Few-shot adaptation performance for PF utility maximization in OOD scenarios.

Fig. 8: Few-shot adaptation performance for the QoS-aware objective in OOD scenarios: (a) average sum rate and (b) average rate of users violating the minimum threshold (

R_{k}<R_{\min}

Fig. 9: Normalized sum rate performance in the OOD scenario

\mathcal{D}_{18}

with varying model sizes.

Fig. 10: Impact of the pre-training masking ratio

\rho

on downstream OOD performance in the

\mathcal{D}_{18}

scenario.

Fig. 11: Ablation study in the

\mathcal{D}_{18}

scenario comparing the proposed hybrid strategy against the edge prediction only variant and the baseline without pre-training.

Wireless Resource Allocation

Wireless resource allocation assigns spectrum, power, beams, and computing capacity to meet throughput, latency, fairness, and energy goals.

See all 174 papers

A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models
Jun 9, 2026Yuxuan Shi, Tingting Yang, Kangning Ma +4Wireless Foundation Models Adaptive Model Routing
Department of Broadband Communication, Pengcheng Laboratory, Shenzhen, 518000, China
Wireless Foundation Models: State-of-the-Art and Open Challenges
Sep 4, 2026Alonso M. Pacheco Huachaca, Juan J. Rodriguez Rodriguez, Ahmed Aboulfotouh +3Wireless Foundation Models Wireless Communications
Institute of Computing, University of Campinas, SP, Brazil · Department of Electrical and Software Engineering, University of Calgary, Calgary, AB, Canada
Graph Signal Diffusion Models for Wireless Resource Allocation
Apr 6, 2026Yigit Berkay Uslu, Samar Hadou, Shirin Saeedi Bidokhti +1Wireless Resource Allocation Diffusion Policy
Dept. of Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, PA, USA

Jun 9, 2026J/K move · Enter open · S save

A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models

Yuxuan Shi, Tingting Yang, Kangning Ma, Liwen Jing+3

Department of Broadband Communication, Pengcheng Laboratory, Shenzhen, 518000, China

Wireless Foundation Models Adaptive Model Routing Multi-Task Learning Parameter-Efficient Fine-Tuning

Open paper AskPDF arXiv

Though wireless foundation models (WFMs) have shown strong potential in learning universal channel representations, their adaptation to various downstream tasks remains constrained by existing paradigms. Fine-tuning strategies introduces substantial computational and storage overhead, while frozen feature extraction leads to sub-optimal performance across diverse downstream tasks. To address this issue, we propose a unified adaptive feature composition framework for multitask generalization in WFMs, where the key component is the Routing Adapter for Feature Composition (RAFC). Instead of extracting only the final-layer output, this router treats the hidden states from different Transformer depths as a reusable pool of multi-level hidden features, and employs a lightweight task-driven feature composition network to generate layer-wise aggregation weights, then adaptively combine hierarchical representations through weighted summation. This design enables each downstream task to access suitable mixture of low-, mid-, and high-level wireless features without modifying the pretrained backbone. Extensive experiments on four representative wireless tasks demonstrate that RAFC consistently outperforms conventional adaptation baselines while introducing fewer than 50K additional parameters. Moreover, the learned routing weights provide interpretable evidence of task-specific layer preferences, making the proposed framework a low-complexity, scalable, and explainable interface for adapting WFMs to diverse downstream scenarios.

Abstract

Wireless Resource Allocation

Explore similar work

A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models

Wireless Foundation Models: State-of-the-Art and Open Challenges

Graph Signal Diffusion Models for Wireless Resource Allocation