A Generic Service-Oriented Function Offloading Framework for Connected Automated Vehicles

TL;DR

This work introduces SOFOF, a generic service-oriented function offloading framework for connected automated vehicles that enables distribution of computational tasks between onboard resources and MEC using modular, pluggable decision-making components. It defines two roles, service provider and service requester, and demonstrates centralized and local offloading decision making within a trajectory planning use case, including AL-ILQR-based planning and ETSI ITS data flows. Through both simulation and real-world experiments, SOFOF demonstrates QoS feasibility and CPU-load reductions on CAVs, while highlighting the system's adaptability to multiple simultaneous offloading requests and dynamic network conditions. The framework is designed to be extensible to additional services and future work will address safety analysis, multi-service orchestration, and scalability of the evaluation setup.

Abstract

Function offloading is a promising solution to address limitations concerning computational capacity and available energy of Connected Automated Vehicles~(CAVs) or other autonomous robots by distributing computational tasks between local and remote computing devices in form of distributed services. This paper presents a generic function offloading framework that can be used to offload an arbitrary set of computational tasks with a focus on autonomous driving. To provide flexibility, the function offloading framework is designed to incorporate different offloading decision making algorithms and quality of service~(QoS) requirements that can be adjusted to different scenarios or the objectives of the CAVs. With a focus on the applicability, we propose an efficient location-based approach, where the decision whether tasks are processed locally or remotely depends on the location of the CAV. We apply the proposed framework on the use case of service-oriented trajectory planning, where we offload the trajectory planning task of CAVs to a Multi-Access Edge Computing~(MEC) server. The evaluation is conducted in both simulation and real-world application. It demonstrates the potential of the function offloading framework to guarantee the QoS for trajectory planning while improving the computational efficiency of the CAVs. Moreover, the simulation results also show the adaptability of the framework to diverse scenarios involving simultaneous offloading requests from multiple CAVs.

Figures (6)

• Figure 1: Considered intersection scenario for service-oriented function offloading including multiple CAVs and a MEC server.
• Figure 2: Overview of SOFOF, simplified for one SP and one SR. After an offloading request is denoted as successful with either the local (LODM) or centralized offloading decision making (CODM), the two service management instances handle the data flow for the offloaded services, respectively. A QoS check on the SR initiates a fallback if the QoS requirements are violated. The CAM check on the SP continuously checks the position of the CAV.
• Figure 3: Data flow between the different components of the function offloading use case service-oriented trajectory planning.
• Figure 4: Real-world street layout with the pilot junction and the two round-trips that were emulated in the simulation in the large picture. The zoomed detailed picture shows the same junction with the routes of the real-world tests. Both pictures use the same color coding for the routes given in the legend. Map from OpenStreetMap.
• Figure 5: Average offloading duration $t_\text{off}$ depending on the maximum inter-arrival time $\Delta t_\text{max}$ with scenarios of varying number of CAVs.