Graph Signal Diffusion Models for Wireless Resource Allocation

2026-04-06

AI Summary

This work addresses the challenge of constrained ergodic resource optimization in wireless networks under graph-structured interference. It introduces, for the first time, a graph signal diffusion model into wireless resource allocation by proposing a conditional diffusion framework based on a graph neural network U-Net architecture. The approach formulates resource allocation as a stochastic signal over graphs and learns the policy distribution generated by a primal-dual expert algorithm, enabling direct sampling of near-optimal solutions during inference without iterative optimization. Experimental results demonstrate that the proposed method achieves near-optimal time-averaged sum rates in power control tasks while effectively satisfying minimum rate constraints, and exhibits strong generalization and transfer capabilities across diverse network topologies and operational states.

Abstract

We consider constrained ergodic resource optimization in wireless networks with graph-structured interference. We train a diffusion model policy to match expert conditional distributions over resource allocations. By leveraging a primal-dual (expert) algorithm, we generate primal iterates that serve as draws from the corresponding expert conditionals for each training network instance. We view the allocations as stochastic graph signals supported on known channel state graphs. We implement the diffusion model architecture as a U-Net hierarchy of graph neural network (GNN) blocks, conditioned on the channel states and additional node states. At inference, the learned generative model amortizes the iterative expert policy by directly sampling allocation vectors from the near-optimal conditional distributions. In a power-control case study, we show that time-sharing the generated power allocations achieves near-optimal ergodic sum-rate utility and near-feasible ergodic minimum-rates, with strong generalization and transferability across network states.

Problem

• wireless resource allocation
• graph signal
• ergodic optimization
• constrained optimization
• interference graph

Innovation

• Graph Signal Diffusion
• Graph Neural Networks
• Wireless Resource Allocation
• Primal-Dual Expert Policy
• Amortized Inference