FedORA: Resource Allocation for Federated Learning in ORAN using Radio Intelligent Controllers

TL;DR

Federated Learning typically struggles under dynamic, multi-RAT network conditions due to communication bottlenecks and energy constraints. FedORA addresses this by integrating FL with ORAN through a two-stage optimization: an RL-based rApp in the Non-RT RIC selects the optimal RAT and allocates power, while a model-based xApp in the Near-Real-Time RIC performs near-real-time resource allocation for FL parameter transmissions. Key contributions include dynamic RAT selection, energy-aware resource management, and improved FL efficiency via ORAN-driven load balancing and privacy-preserving aggregation. On CIFAR-10 with CNNs, FedORA achieves competitive accuracy and lower power consumption compared with FedAvg, FLAIR, and Greedy FL under dynamic network conditions. The work demonstrates a practical, scalable pathway for real-time, energy-efficient, privacy-preserving collaborative learning in multi-RAT ORAN deployments.

Abstract

This work proposes an integrated approach for optimising Federated Learning (FL) communication in dynamic and heterogeneous network environments. Leveraging the modular flexibility of the Open Radio Access Network (ORAN) architecture and multiple Radio Access Technologies (RATs), we aim to enhance data transmission efficiency and mitigate client-server communication constraints within the FL framework. Our system employs a two-stage optimisation strategy using ORAN's rApps and xApps. In the first stage, Reinforcement Learning (RL) based rApp is used to dynamically select each user's optimal Radio Access Technology (RAT), balancing energy efficiency with network performance. In the second stage, a model-based xApp facilitates near-real-time resource allocation optimisation through predefined policies to achieve optimal network performance. The dynamic RAT selection and resource allocation capabilities enabled by ORAN and multi-RAT contribute to robust communication resilience in dynamic network environments. Our approach demonstrates competitive performance with low power consumption compared to other state-of-the-art models, showcasing its potential for real-time applications demanding both accuracy and efficiency. This robust and comprehensive framework, enabling clients to utilise available resources effectively, highlights the potential for scalable, collaborative learning applications prioritising energy efficiency and network performance.

Figures (2)

• Figure 1: The proposed FedORA Network layout
• Figure 2: Comparative performance of our model's accuracy and loss against other relevant models.