Dynamic Adaptive Federated Learning for mmWave Sector Selection

Beamforming techniques use massive antenna arrays to formulate narrow Line-of-Sight signal sectors to address the increased signal attenuation in millimeter Wave (mmWave). However, traditional sector selection schemes involve extensive searches for the highest signal strength sector, introducing extra latency and communication overhead. This paper introduces a dynamic layer-wise and clustering-based federated learning (FL) algorithm for beam sector selection in autonomous vehicle networks called enhanced Dynamic Adaptive FL (eDAFL). The algorithm detects and selects the most important layers of a machine learning model for aggregation in FL process, significantly reducing network overhead and failure risks. eDAFL also consider an intra-cluster and inter-cluster approach to reduce overfitting and increase the abstraction level. We evaluate eDAFL on a real-world multi-modal dataset, demonstrating improved model accuracy by approximately 6.76% compared to existing methods, while reducing inference time by 84.04% and model size up to 52.20%.

Dynamic Adaptive Federated Learning for mmWave Sector Selection

Lucas Pacheco, Federal University of Pará; Torsten Braun, University of Bern; Kaushik Chowdhury, Northeastern University, USA; Denis Rosario, Federal University of Pará (UFPA); Batool Salehi, Northeastern University; Eduardo Cerqueira, Federal University of Para