logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

Spatiotemporal-aware task offloading with backhaul optimization for vehicular edge computing

delete2026-02-27
delete0
PRE
AI
A
Aoran Li
H
Honglong Chen *
Z
Zhishuai Li
N
Ning Chen
Z
Zhichen Ni
H
Haiyang Sun
DOI：10.1016/j.comcom.2026.108476delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
Intelligent Transportation Systems (ITS) generate increasing volumes of vehicle data, creating unprecedented computational challenges. Although cloud computing offers abundant computational resources, its reliance on distant data centers often introduces delay and bandwidth challenges for real-time ITS processing. Vehicular edge computing solves this by distributing resources to the network edge. This enables efficient task offloading to nearby Intelligent Connected Vehicles, significantly reducing delays while improving responsiveness and resource utilization. However, as task data volumes grow, the backhaul phase becomes a dominant bottleneck in resource-constrained traffic hotspot environments, causing existing offloading methods to struggle with maintaining efficient task processing. The key contribution of this work is a collaborative offloading decision model that optimizes task allocation in vehicular networks by addressing both delay and resource constraints. To solve the problem, we propose three principal contributions: First, we develop a collaborative offloading decision model based on multi-hop communication, which generates optimal offloading node selection strategies for each task. Second, we formulate the vehicular task offloading problem as a Flexible Job Shop Scheduling Problem with Uncertain Processing Times. Finally, we design the Dijkstra Genetic Algorithm (DGA), a novel hybrid optimization method that integrates the Dijkstra multi-strategy approach with genetic algorithms to enhance offloading efficiency. Simulation results show significant performance gains, including a 12% reduction in the optimization objective and notable improvements in overall fitness compared with existing approaches.
Keywords:
Vehicular edge computing
Task offloading
Backhaul optimization
Multi-hop communication
Flexible Job Shop Scheduling

Journal

Computer Communications cover
Computer Communications
IF:
4.3
Papers:
544
Citations:
1.1W

Organization

No organization information available