Joint Partial Offloading and Resource Allocation for Vehicular Federated Learning Tasks

In the foreseeable Intelligent Transportation System, Intelligent Connected Vehicles (ICVs) will play an important role in improving travel efficiency and safety. However, it is challenging for ICVs to support the resource-hungry autonomous driving applications due to the limitation of hardware computing power. Fortunately, the emergence of Multi-access Edge Computing helps overcome this limitation effectively. This paper addresses the vehicle-to-edge server computation offloading conundrum by optimizing the trade-offs in partial offloading and resource allocation. Proposing a distributed approach, this study confronts the multi-variable non-convex challenge directly by decoupling variables and deriving constraint-based bounds that guide the decisions for offloading and allocation. A novel low-complexity distributed algorithm is introduced that not only tends toward optimal but also demonstrates superior real-time applicability and efficiency, illustrated through enhanced performances both in simulated trials and genuine vehicular edge computing settings. The algorithm’s practical effectiveness addresses a notable gap between the theoretical models for computation offloading and actual real-life execution, reinforcing the soundness and relevance of the proposed method. Furthermore, its advanced integration with federated learning frameworks marks a leading-edge application, substantiating significant enhancements in computational efficiency and robustness.

I. Introduction

The evolution towards autonomous driving generates an abundance of data, requiring efficient real-time processing capabilities for advanced applications such as image processing and dynamic path planning [1], [2], [3], [4], [5]. To adequately meet these demands, Multi-access Edge Computing (MEC) steps in to offer proximity-based computational resources, allowing ICVs to offload these intensive workflows and mitigate latency issues [6], [7], [8]. Alongside, Federated Learning (FL) surfaces as a distributed framework enabling collaborative model training across vehicular networks. The combination of these two frameworks under the computational offloading paradigm substantially elevates the efficiency of utilizing the wealth of data from ICVs for training generalized models, thereby enhancing the operational capabilities of autonomous driving technologies [9], [10], [11], [12].