[Paper Review] V2V-Based Task Offloading and Resource Allocation in Vehicular Edge Computing Networks

Junjin He, Yujie Wang+2 more|arXiv (Cornell University)|Dec 30, 2021

Vehicular Ad Hoc Networks (VANETs)4 citations

TL;DR

This paper proposes OPFTO, a V2V-based task offloading and resource allocation framework that leverages idle computing resources in mobile vehicles to reduce latency when RSUs are unavailable. By integrating a mobility-aware pre-filtering mechanism with an improved Hybrid Gravitational Search Algorithm (HGSA), the system achieves lower task completion delay and higher success rates than baseline strategies, especially under dynamic vehicular conditions.

ABSTRACT

In the research and application of vehicle ad hoc networks (VANETs), it is often assumed that vehicles obtain cloud computing services by accessing to roadside units (RSUs). However, due to the problems of insufficient construction quantity, limited communication range and overload of calculation load of roadside units, the calculation mode relying only on vehicle to roadside units is difficult to deal with complex and changeable calculation tasks. In this paper, when the roadside unit is missing, the vehicle mobile unit is regarded as a natural edge computing node to make full use of the excess computing power of mobile vehicles and perform the offloading task of surrounding mobile vehicles in time. In this paper, the OPFTO framework is designed, an improved task allocation algorithm HGSA is proposed, and the pre-filtering process is designed with full consideration of the moving characteristics of vehicles. In addition, vehicle simulation experiments show that the proposed strategy has the advantages of low delay and high accuracy compared with other task scheduling strategies, which provides a reference scheme for the construction of Urban Intelligent Transportation in the future.

Motivation & Objective

• Address the limitations of RSU-dependent task offloading in VANETs due to sparse coverage, high load, and infrastructure costs.
• Utilize idle computing resources in mobile vehicles as distributed edge nodes to enhance system resilience and efficiency.
• Design a robust, real-time task offloading system that accounts for vehicle mobility and dynamic network topology.
• Minimize task completion delay and improve resource utilization through intelligent candidate selection and task allocation.
• Develop a mobility-aware algorithm that maintains performance stability regardless of vehicle location or stay duration.

Proposed method

• Propose OPFTO, an online pre-filtering task offloading framework with two stages: candidate vehicle selection and task allocation.
• Introduce a pre-filtering process that evaluates vehicle stay time, task completion time, and historical offloading success rate to select reliable candidates.
• Design the HGSA algorithm—a hybrid metaheuristic based on simulated annealing—optimized for minimizing task completion delay.
• Incorporate mobility constraints into the optimization process by ensuring selected vehicles can complete assigned tasks within their stay duration.
• Use a multi-objective fitness function in HGSA that balances load distribution and minimizes the maximum completion time across vehicles.
• Simulate V2V task offloading in urban scenarios with varying numbers of vehicles and subtasks to evaluate performance.

Experimental results

Research questions

• RQ1 How can task offloading be effectively performed in VANETs when roadside units are unavailable or overloaded?
• RQ2 What role does vehicle mobility play in determining the success and efficiency of V2V task offloading?
• RQ3 How can candidate vehicles be pre-filtered to improve offloading success rate and reduce delay?
• RQ4 To what extent does the HGSA algorithm outperform traditional offloading strategies in terms of delay and resource utilization?
• RQ5 How robust is the proposed system to variations in vehicle stay duration and network dynamics?

Key findings

• The HGSA algorithm consistently achieves the lowest task completion delay across all tested configurations, outperforming RANDOM, SCF, DF, and SSF, especially when the number of subtasks is small.
• When the number of subtasks increases, HGSA maintains low delay by distributing tasks across all available vehicles, avoiding bottlenecks caused by long-stay vehicles.
• The pre-filtering mechanism significantly improves offloading success rate by ensuring candidate vehicles can complete assigned tasks within their stay duration.
• HGSA demonstrates strong robustness to vehicle mobility, maintaining stable performance regardless of task vehicle location or stay time, unlike SCF, DF, and SSF, which are highly sensitive to vehicle position.
• The inclusion of candidate selection reduces re-offloading overhead caused by vehicle departure, as confirmed by a 20% reduction in delay when candidate filtering is applied.
• In scenarios with 20 vehicles and 1400 subtasks, HGSA reduced total completion delay by up to 35% compared to the next-best baseline, SSF.