Modeling Edge-to-Cloud Offloading Workloads for Autonomous Vehicles

TL;DR

This paper introduces a workload modeling framework for autonomous vehicle data offloading, capturing unique data characteristics and analyzing real-world mobility traces to inform edge-to-cloud offloading strategies.

Contribution

It presents a novel system-level workload model for autonomous vehicle data, incorporating empirical data and classifying data types for more accurate offloading analysis.

Findings

Workload scales with vehicle penetration.

Workload shows temporal structure and spatial imbalance.

Distinguished from baseline traffic models.

Abstract

Autonomous vehicles generate large volumes of data for applications such as fleet monitoring, model retraining, and high-definition map updates. Existing studies often rely on generic traffic traces, which do not capture the characteristics of autonomous driving workloads. This paper proposes a system-level workload modeling framework for vehicle-to-cloud data. We classify offloaded data into three types: telemetry, event-driven fleet learning, and high-definition map updates, while we model their generation using a parameterized formulation based on empirical data. Using a real-world mobility trace from Munich, we analyze the resulting workloads over time and space. The results show that workload scales with vehicle penetration, exhibits temporal structure and spatial imbalance across access points, and is distinguished from baseline traffic models.