AIGC-assisted Federated Learning for Vehicular Edge Intelligence: Vehicle Selection, Resource Allocation and Model Augmentation

2025-03-25

AI Summary

In vehicular edge intelligence, high vehicle mobility, unstable wireless channels, and data heterogeneity severely impede federated learning (FL) performance—causing slow convergence, low accuracy, and high end-to-end latency. Method: This paper proposes GenFV, an AIGC-enhanced FL framework tailored for vehicular networks. It pioneers the use of Artificial Intelligence-Generated Content (AIGC) to synthesize Non-IID training data, mitigating data scarcity and skew. We introduce an Earth Mover’s Distance (EMD)-based weighted heterogeneity metric and provide theoretical convergence analysis. Furthermore, we design a bi-scale Mixed-Integer Nonlinear Programming (MINLP) optimizer integrating velocity-aware vehicle selection, label sharing, and joint allocation of bandwidth, transmit power, and synthetic data. Results: Experiments under dynamic, resource-constrained vehicular environments demonstrate that GenFV significantly improves model accuracy and convergence robustness while reducing system end-to-end latency by over 35%, outperforming state-of-the-art baselines across multiple metrics.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Distributed Machine Learning & Federated Learning
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

To leverage the vast amounts of onboard data while ensuring privacy and security, federated learning (FL) is emerging as a promising technology for supporting a wide range of vehicular applications. Although FL has great potential to improve the architecture of intelligent vehicular networks, challenges arise due to vehicle mobility, wireless channel instability, and data heterogeneity. To mitigate the issue of heterogeneous data across vehicles, artificial intelligence-generated content (AIGC) can be employed as an innovative data synthesis technique to enhance FL model performance. In this paper, we propose AIGC-assisted Federated Learning for Vehicular Edge Intelligence (GenFV). We further propose a weighted policy using the Earth Mover's Distance (EMD) to quantify data distribution heterogeneity and introduce a convergence analysis for GenFV. Subsequently, we analyze system delay and formulate a mixed-integer nonlinear programming (MINLP) problem to minimize system delay. To solve this MINLP NP-hard problem, we propose a two-scale algorithm. At large communication scale, we implement label sharing and vehicle selection based on velocity and data heterogeneity. At the small computation scale, we optimally allocate bandwidth, transmission power and amount of generated data. Extensive experiments show that GenFV significantly improves the performance and robustness of FL in dynamic, resource-constrained environments, outperforming other schemes and confirming the effectiveness of our approach.

Problem

• Addressing data heterogeneity in vehicular federated learning using AIGC.
• Optimizing system delay via mixed-integer nonlinear programming in edge networks.
• Enhancing FL robustness with dynamic vehicle and resource selection.

Innovation

• AIGC-assisted FL for data heterogeneity mitigation
• EMD-based weighted policy for data distribution
• Two-scale algorithm for MINLP problem solving