Dependency-Aware Task Offloading in Cooperative UAV-HAPS-Assisted Vehicular Networks

The advent of Intelligent Transportation Systems (ITS) stimulated the deployment of connected and autonomous vehicles supporting computation-intensive and delay-sensitive applications. This evolution has underscored the challenges of user device's limited resources in vehicular networks. Addressing them requires innovative schemes, particularly in the context of integrated Unmanned Aerial Vehicles (UAVs) and High-Altitude Platform Stations (HAPS) networks. In this paper, we propose a novel UAV/HAPS-enabled vehicular framework aiming to enhance task management. It is predicated on a collaborative offloading scheme that caters to the Quality-of-Service requirements of multi-vehicular tasks and aims to minimize the overall energy consumption within the dynamic ITS environment. Moreover, the proposed scheme is designed to handle the decomposition of tasks into interdependent sub-tasks. Given the NP-hardness of the offloading problem, we develop an approach based on the Genetic Algorithm. Through simulations, we demonstrate the superiority of our method compared to benchmarks, in terms of successful service rate and energy consumption. Our method constitutes a substantial advancement in handling ITS services through a cooperative UAV-HAPS framework.