Dynamic Scheduling for Vehicle-to-Vehicle Communications Enhanced Federated Learning

2024-06-25

IEEE Transactions on Wireless Communications

AI Summary

To address the degradation of vehicle-mounted federated learning (VFL) training performance caused by high vehicular mobility and energy constraints in vehicular networks, this paper proposes a V2V communication-enhanced dynamic scheduling framework. Methodologically, it introduces a novel derivative-based drift-penalty approach that reformulates long-term stochastic optimization as an online-solvable mixed-integer nonlinear programming (MINLP) problem; further, a scheduling priority decomposition strategy is proposed to decouple the original problem into a set of efficiently solvable convex subproblems. The framework integrates V2V protocol modeling, onboard federated learning architecture, and interior-point method-based optimization. Experimental evaluation on CIFAR-10 image classification achieves a 3.18% accuracy improvement, while Argoverse trajectory prediction yields a 10.21% reduction in average displacement error—both significantly outperforming state-of-the-art baselines.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Planning, Routing, and Scheduling: Planning/Scheduling and Learning
• Machine Learning: Distributed Machine Learning & Federated Learning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Leveraging the computing and sensing capabilities of vehicles, vehicular federated learning (VFL) has been applied to edge training for connected vehicles. The dynamic and interconnected nature of vehicular networks presents unique opportunities to harness direct vehicle-to-vehicle (V2V) communications, enhancing VFL training efficiency. In this paper, we formulate a stochastic optimization problem to optimize the VFL training performance, considering the energy constraints and mobility of vehicles, and propose a V2V-enhanced dynamic scheduling (VEDS) algorithm to solve it. The model aggregation requirements of VFL and the limited transmission time due to mobility result in a stepwise objective function, which presents challenges in solving the problem. We thus propose a derivative-based drift-plus-penalty method to convert the long-term stochastic optimization problem to an online mixed integer nonlinear programming (MINLP) problem, and provide a theoretical analysis to bound the performance gap between the online solution and the offline optimal solution. Further analysis of the scheduling priority reduces the original problem into a set of convex optimization problems, which are efficiently solved using the interior-point method. Experimental results demonstrate that compared with the state-of-the-art benchmarks, the proposed algorithm enhances the image classification accuracy on the CIFAR-10 dataset by 3.18% and reduces the average displacement errors on the Argoverse trajectory prediction dataset by 10.21%.

Problem

• Optimizing VFL training with V2V communications under energy constraints
• Solving stepwise objective challenges in dynamic vehicular networks
• Enhancing accuracy and reducing errors in vehicle learning tasks

Innovation

• Dynamic scheduling algorithm for V2V-enhanced VFL
• Derivative-based drift-plus-penalty for stochastic optimization
• Interior-point method for convex optimization problems