Distributed Learning for Reliable and Timely Communication in 6G Industrial Subnetworks

TL;DR

This paper introduces a distributed, learning-based protocol for 6G industrial subnetworks that enhances timely communication by reducing collisions and adapting to dynamic conditions, using lightweight neural models and online training.

Contribution

It proposes a novel distributed learning approach with implicit inter-subnetwork coordination for collision-aware access in industrial 6G networks, suitable for constrained environments.

Findings

Significantly improves timely packet delivery probability.

Achieves 21% gain over classical Multi-Armed Bandit methods.

Effective in dense, high-load industrial scenarios.

Abstract

Emerging 6G industrial networks envision autonomous in-X subnetworks to support efficient and cost-effective short range, localized connectivity for autonomous control operations. Supporting timely transmission of event-driven, critical control traffic is challenging in such networks is challenging due to limited radio resources, dynamic device activity, and high mobility. In this paper, we propose a distributed, learning-based random access protocol that establishes implicit inter-subnetwork coordination to minimize the collision probability and improves timely delivery. Each subnetwork independently learns and selects access configurations based on a contention signature signal broadcast by a central access point, enabling adaptive, collision-aware access under dynamic traffic and mobility conditions. The proposed approach features lightweight neural models and online training, making it suitable for deployment in constrained industrial subnetworks. Simulation results show that our method significantly improves the probability of timely packet delivery compared to baseline methods, particularly in dense and high-load scenarios. For instance, our proposed method achieves 21% gain in the probability of timely packet delivery compared to a classical Multi-Armed Bandit (MAB) for an industrial setting of 60 subnetworks and 5 radio channels.