CHOMET: Conditional Handovers via Meta-Learning

Handovers (HOs) are the cornerstone of modern cellular networks for enabling seamless connectivity to a vast and diverse number of mobile users. However, as mobile networks become more complex with more diverse users and smaller cells, traditional HOs face significant challenges, such as prolonged delays and increased failures. To mitigate these issues, 3GPP introduced conditional handovers (CHOs), a new type of HO that enables the preparation (i.e., resource allocation) of multiple cells for a single user to increase the chance of HO success and decrease the delays in the procedure. Despite its advantages, CHO introduces new challenges that must be addressed, including efficient resource allocation and managing signaling/communication overhead from frequent cell preparations and releases. This paper presents a novel framework aligned with the O-RAN paradigm that leverages meta-learning for CHO optimization, providing robust dynamic regret guarantees and demonstrating at least 180% superior performance than other 3GPP benchmarks in volatile signal conditions.

On the Dynamic Regret of Following the Regularized Leader: Optimism with History Pruning

AZTEC+: Long and Short Term Resource Provisioning for Zero-Touch Network Management

In the past few years, network infrastructures have transitioned from prominently hardware-based models to networks of functions, where software components provide the required functionalities with unprecedented scalability and flexibility. However, this new vision entails a completely new set of problems related to resource provisioning and the network function operation, making it difficult to manage the network function lifecycle management with traditional, human-in-the-loop approaches. Novel zero-touch management solutions promise autonomous network operation with limited human interactions. However, modeling network function behavior into compelling variables and algorithm is an aspect that such solutions must take into account. In this paper, we propose AZTEC+, a data-driven solution for anticipatory resource provisioning in network slicing scenarios. By leveraging a hybrid and modular deep learning architecture, AZTEC+ not only forecasts the future demands for target services but also identifies the best trade-offs to balance the costs due to the instantiation and reconfiguration of such resources. Our experimental evaluation, based on real-world network data, shows how AZTEC+ can outperform state-of-the-art management solutions for a large set of metrics.

Devoxx France – 13ème édition – du 16 au 18 avril 2025

Demonstrating Distributed Inference in the User Plane with DUNE

Deploying Machine Learning (ML) models in the user plane enables low-latency and scalable in-network inference, but integrating them into programmable devices faces stringent constraints in terms of memory resources and computing capabilities. In this demo, we show how the newly proposed DUNE, a novel framework for distributed user-plane inference across multiple programmable network devices by automating the decomposition of large ML models into smaller sub-models, mitigates the limitations of traditional monolithic ML designs.We run experiments on a testbed with Intel Tofino switches using measurement data and show how DUNE not only improves the accuracy that the traditional single-device monolithic approach gets but also maintains a comparable per-switch latency.