Anti-Byzantine Attacks Enabled Vehicle Selection for Asynchronous Federated Learning in Vehicular Edge Computing

TL;DR

This work addresses AFL in vehicular edge computing by jointly considering vehicle mobility, data volume, computation capabilities, and time-varying channels, along with Byzantine attacks. It proposes a DRL framework using DDPG to select participating vehicles and to drive a latency- and loss-aware aggregation scheme, including a weighted update and a threshold-based Byzantine defense using a trusted RSU dataset. The method demonstrates improved global model accuracy and robustness against attacks compared to traditional FL and AFL, with convergence of the DRL policy and tangible gains in test performance. The approach offers a practical, attack-resilient pathway to safer and more reliable AFL in dynamic VEC environments.

Abstract

In vehicle edge computing (VEC), asynchronous federated learning (AFL) is used, where the edge receives a local model and updates the global model, effectively reducing the global aggregation latency.Due to different amounts of local data,computing capabilities and locations of the vehicles, renewing the global model with same weight is inappropriate.The above factors will affect the local calculation time and upload time of the local model, and the vehicle may also be affected by Byzantine attacks, leading to the deterioration of the vehicle data. However, based on deep reinforcement learning (DRL), we can consider these factors comprehensively to eliminate vehicles with poor performance as much as possible and exclude vehicles that have suffered Byzantine attacks before AFL. At the same time, when aggregating AFL, we can focus on those vehicles with better performance to improve the accuracy and safety of the system. In this paper, we proposed a vehicle selection scheme based on DRL in VEC. In this scheme, vehicle s mobility, channel conditions with temporal variations, computational resources with temporal variations, different data amount, transmission channel status of vehicles as well as Byzantine attacks were taken into account.Simulation results show that the proposed scheme effectively improves the safety and accuracy of the global model.

Figures (12)

• Figure 1: System model
• Figure 2: DDAFL flow diagram
• Figure 3: Epoch reward in training stage
• Figure 4: Loss in testing stage with bad node
• Figure 5: Loss in testing stage without bad node