Mobility and dependency-aware task offloading for intelligent assisted driving in vehicular edge computing networks

Abstract

Intelligent assisted driving is an important application in vehicular edge computing networks (VECNs). In the intelligent transportation system (ITS), a group of moving vehicle users need to be coordinated to complete complex vehicular applications. A number of dependent, latency-sensitive, and computation-intensive tasks are generated. However, the existing works have given less consideration to the dependencies among both vehicle users and the subtasks in vehicle, which makes it a huge challenge to complete tasks timely. When interdependent tasks come from different vehicle users, a special task preparation time is needed, which can disrupt the ongoing task processing. Furthermore, the high mobility of vehicles directly affects the data transmission rate. To address the mentioned challenges, we design an efficient mobility and dependency-aware task offloading strategy in VECNs. The objective is to minimize both the overall system task completion delay and the economic cost. We take into account the real-time locations and task preparation time of vehicle users. Additionally, we propose a multi-decision-making offloading algorithm (MDOA) that primarily analyzes the processing priorities for both vehicle users and subtasks. In order to integrate practical applications, the financial expenses of vehicle users are also considered as an indispensable part. As a result, we propose an efficient two-step task offloading algorithm. Through numerous simulation examples, we demonstrate the efficiency and high performance of the proposed task offloading strategies in VECNs when compared to existing algorithms.