NetDiff: Deep Graph Denoising Diffusion for Ad Hoc Network Topology Generation

Félix Marcoccia, Gilles Monzat de Saint Julien, Cédric Adjih, Paul Mühlethaler

Keywords: graph diffusion, wireless networks, network topology

TL;DR: We present NetDiff, a diffusion-based solution with architectural augmentations that generates high-performance wireless network topologies.

Abstract: This paper introduces NetDiff, an expressive graph denoising diffusion probabilistic architecture that generates high-performance link topologies for wireless ad hoc networks. Such networks, with directional antennas, can achieve unmatched throughput and scalability when the communication links are designed to provide good geometric properties, notably by reducing interference between these links while respecting diverse physical constraints. How to craft such a link assignment algorithm remains a real problem. Deep graph generation offers multiple advantages compared to traditional approaches: it allows to relieve the network nodes of the communication burden caused by the search of viable links and to avoid resorting to heavy combinatorial methods to find a good link topology. Given that graph neural networks sometimes tend to struggle with global, structural properties, we augment the popular graph transformer with cross-attentive modulation tokens in order to improve global control over the predicted topology. We also incorporate simple node and edge features, as well as additional loss terms, to facilitate the compliance with the network topology physical constraints. A network evolution algorithm based on partial diffusion is proposed to maintain the network topology over time when the nodes are moving. Our results show that the generated topologies are realistic, require only minor correction steps to be operational, and establish NetDiff as a viable solution for maximizing the benefits offered by directional antennas.

Submission Number: 2