A Survey on UAV-enabled Edge Computing Resource Management Perspective

2025-04-24 0 0 1.01MB 36 页 10玖币

侵权投诉

A Survey on UAV-enabled Edge Computing: Resource

Management Perspective

XIAOYU XIA, School of Computing Technologies, RMIT University, Australia

SHEIK MOHAMMAD MOSTAKIM FATTAH, Centre for Research on Engineering Software Tech-

nologies, University of Adelaide, Australia

MUHAMMAD ALI BABAR, Centre for Research on Engineering Software Technologies, University of

Adelaide, Australia

Edge computing facilitates low-latency services at the network’s edge by distributing computation, communi-

cation, and storage resources within the geographic proximity of mobile and Internet-of-Things (IoT) devices.

The recent advancement in Unmanned Aerial Vehicles (UAV) technologies has opened new opportunities

for edge computing in military operations, disaster response, or remote areas where traditional terrestrial

networks are limited or unavailable. In such environments, UAVs can be deployed as aerial edge servers or

relays to facilitate edge computing services. This form of computing is also known as UAV-enabled Edge

Computing (UEC), which oers several unique benets such as mobility, line-of-sight, exibility, computational

capability, and cost-eciency. However, the resources on UAVs, edge servers, and IoT devices are typically

very limited in the context of UEC. Ecient resource management is, therefore, a critical research challenge

in UEC. In this article, we present a survey on the existing research in UEC from the resource management

perspective. We identify a conceptual architecture, dierent types of collaborations, wireless communication

models, research directions, key techniques and performance indicators for resource management in UEC. We

also present a taxonomy of resource management in UEC. Finally, we identify and discuss some open research

challenges that can stimulate future research directions for resource management in UEC.

CCS Concepts: •General and reference

→

Surveys and overviews;•Computer systems organization

→Cloud computing.

Additional Key Words and Phrases: UAV-enabled edge computing, resource management, architectures,

ooading, allocation, provisioning, algorithms

ACM Reference Format:

Xiaoyu Xia, Sheik Mohammad Mostakim Fattah, and Muhammad Ali Babar. 2024. A Survey on UAV-enabled

Edge Computing: Resource Management Perspective. ACM Comput. Surv. 1, 1, Article 1 (January 2024),

36 pages. https://doi.org/XXX/XXXX

1 INTRODUCTION

1.1 Edge Computing and UAVs

The recent advancements of the Internet of Things (IoT) and wireless communication technologies

have introduced many new applications that require high computational power and low latency [

Authors’ addresses: Xiaoyu Xia, School of Computing Technologies, RMIT University, VIC, 3000, Australia, xiaoyu.xia@

rmit.edu.au; Sheik Mohammad Mostakim Fattah, Centre for Research on Engineering Software Technologies, University of

Adelaide, SA, 5005, Australia, sheik.fattah@adelaide.edu.au; Muhammad Ali Babar, Centre for Research on Engineering

Software Technologies, University of Adelaide, SA, 5005, Australia, ali.babar@adelaide.edu.au.

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee

provided that copies are not made or distributed for prot or commercial advantage and that copies bear this notice and

the full citation on the rst page. Copyrights for components of this work owned by others than ACM must be honored.

Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires

prior specic permission and/or a fee. Request permissions from permissions@acm.org.

0360-0300/2024/1-ART1 $XXX

https://doi.org/XXX/XXXX

ACM Comput. Surv., Vol. 1, No. 1, Article 1. Publication date: January 2024.

arXiv:2210.06679v2 [cs.DC] 26 Sep 2023

1:2 Xiaoyu Xia, Sheik Mohammad Mostakim Faah, and Muhammad Ali Babar

including wearable cognitive assistance, augmented reality (AR), smart healthcare, facial recognition,

and road safety monitoring [

171

]. However, IoT devices typically have limited computational

resources, storage, network coverage, and energy. Therefore, resource-intensive IoT applications

often face signicant challenges in maintaining the expected Quality of Services (QoS) [

A conventional approach to maintain the expected QoS is to leverage cloud computing [

] by

utilizing resources from remote cloud servers in dierent forms such as virtual machines, virtual

storage, and virtual private networks [

]. However, cloud computing is now considered inadequate

to meet the low-latency requirements of resource-intensive and delay-sensitive IoT applications

[

]. The reason is two-fold. First, the number of IoT devices is exponentially increasing and it is

expected that it will be approximately 125 billion by 2030 [

]. These devices generate a large volume

of network trac that burdens the backhaul network and signicantly aects its performance by

network congestion [

138

]. Second, cloud servers are typically placed at a remote distance from

IoT devices. As a result, cloud computing introduces a considerable amount of delay in service

provisioning, which degrades the overall QoS of delay-sensitive IoT applications [73, 154].

Edge computing is a relatively new paradigm that oers an alternative computing solution for

delay-sensitive and resource-intensive IoT applications. Edge computing extends cloud computing

technologies to the edge of a network, closer to users and IoT devices [

]. It allows a resource-

limited IoT device (a.k.a., edge devices) to fully or partially ooad its data and computational tasks

to nearby powerful edge servers or other edge devices [

]. It substantially improves the latency and

energy eciency of IoT applications. This also reduces the trac congestion at the core network.

Edge servers also work as data caches to store frequently accessed data by IoT devices to improve

the QoS of the applications [171, 177].

IoT devices are typically connected to an edge infrastructure using wireless networks [

However, a good wireless network infrastructure may not always be available in some of the

remotest areas, e.g., rural or mountain [

]. Moreover, a wireless network infrastructure can easily

be aected by natural disasters such as earthquakes, oods, or storms. In military operations

or emergency rescue missions, it may often be dicult to have a trustworthy wireless network

infrastructure [

]. Unmanned Aerial Vehicle (UAV) technologies have opened a new opportunity

where edge computing services are provisioned using UAVs in military operations, disaster response,

or rural areas. This is also known as UAV-enabled edge computing (UEC) [

]. UAVs oer a wide

range of bets such as mobility, exibility, and cost-eciency, which make UEC a promising

solution. A UAV typically works as an aerial edge server or a relay in a UEC environment [

]. IoT

devices ooad their computational tasks fully or partially to a nearby UAV. A UAV either processes

the tasks locally or sends them for remote execution to nearby edge/cloud servers.

1.2 Resource Management in Edge Computing and UEC

Resource management in edge computing has been investigated intensively by many researchers

in the last decades [

139

]. Edge computing oers many unique advantages, such as distributed

frameworks and low-latency services, however, it poses many unique challenges, including resource

provisioning, computation ooading, and resource allocation.

In general, resource management refers to the use of a set of actions and methodologies to allocate

resources, such as bandwidth, energy, CPU, GPU and many others, to devices or tasks for achieving

specic objectives, i.e., maximizing utility, minimizing service latency, etc [

143

]. In

edge computing, the resource has its new characteristics: 1) limited - the resources on an edge server

are usually constrained due to the limited physical size with smaller processors and a limited power

budget [

]; 2) dynamic - task requirements, device mobility, available resource amounts on

edge servers are always changing over time [

145

]; and 3) heterogeneous - both resources on edge

servers and resources requirements of devices and tasks are heterogeneous [98].

ACM Comput. Surv., Vol. 1, No. 1, Article 1. Publication date: January 2024.

A Survey on UAV-enabled Edge Computing: Resource Management Perspective 1:3

Table 1. Number of related papers

Year 2017 2018 2019 2020 2021 2022

Number

2 14 28 58 67 96

Edge computing provides storage, computing and communication resources at the network

edge to improve serviceability signicantly. To ensure the utility of edge resources, appropriate

resource scheduling strategies are required from two perspectives: infrastructure providers and

services providers. Such resource management problems have been intensively investigated from

the perspective of infrastructure providers with dierent optimization objectives, e.g., minimum

transmission cost [

121

], maximum system throughput [

], maximum data sharing eciency [

etc. From a service provider’s perspective, the objectives of resource management can be minimizing

resource rent cost [

142

], maximizing users’ QoS/QoE [

], maximizing users’ total data rate [

141

maximizing its revenue [103], and many others.

Resource management in UEC is a topical research area, attracting many researchers from

both academia and industry. [

174

]. Table 1 depicts the number of papers related to resource

management in UEC published in the last 5 years. The details of the literature methodology are

explained in Section 1.4. It clearly shows an increasing interest in the research community’s

inecient resource management in UEC. Compared with traditional edge computing, UEC oers

several advantages such as exibility and autonomy with the fast deployment of UAVs for serving

devices in rural areas or for critical missions [

135

136

]. However, UEC also poses many new

challenges with its unique characteristics, such as multiple roles of UAVs, the mobility of servers,

and the UAV trajectory design. Specically, UAVs can play multiple roles when formulating resource

management strategies in UEC, including servers, relays and users. The high mobility and limited

resources of UAVs also complicate resource management in UEC. In addition, external factors, such

as wind, rain and storm, can impact the performance of UAVs signicantly. Considering those

complicated scenarios in UEC, researchers from academia and industry have started to investigate

to identify and develop suitable strategies for resource management in UEC. To guide researchers

in this area, we have conducted this survey to provide an overview of resource management in

UEC and point out the challenges that can direct future research.

1.3 Related Surveys

Existing studies have focused on dierent aspects of resource management in UEC including

resource provisioning and allocation, task or data ooading, and trajectory design. However, there

is no existing survey or review that provides a comprehensive overview of resource management in

UEC. A few survey studies of UEC are mainly limited to the computational ooading aspect of

resource management in UEC. A survey on ooading in UEC is presented in [

]. This study

reports a comparative assessment of various ooading algorithms based on their performance and

features. The survey provides dierent application scenarios and a case study of where UAVs could

be leveraged to enable edge computing. Examples of such applications are next-generation wireless

networking, surveillance of property, UAV-enabled target tracking, UAV-MEC in a pandemic, and

reconnaissance in military operations. It provides a classication of ooading algorithms based on

ooading policies such as binary ooading, partial ooading, and relay.

A short survey is reported in [

] which provides an overview of UEC networks and focuses

on energy eciency-related studies. It summarizes key computational ooading techniques con-

cerning the ground mobile users who ooad computational tasks to nearby UAVs. It also provides

a summary of the approaches for the resource management of UAVs from the energy eciency

ACM Comput. Surv., Vol. 1, No. 1, Article 1. Publication date: January 2024.

1:4 Xiaoyu Xia, Sheik Mohammad Mostakim Faah, and Muhammad Ali Babar

Table 2. A list of recent surveys in UAV-enabled Edge Computing

Paper

Title

Year

Key Contributions

[

174

]

Mobile Edge Computing in Un-

manned Aerial Vehicle Networks

2019

The paper presents three UEC architectures based on the role of

UAVs in UEC. It provides a brief survey on computational ooad-

ing and resource allocation in UEC.

[3]

Energy Ecient UAV-Enabled Mobile

Edge Computing for IoT Devices: A

Review

2021

The paper presents a brief survey on UEC networks with a fo-

cus on energy eciency. It introduces basic terminologies and

architectures used in UEC. It also presents various techniques

and challenges related to computational ooading and a brief

overview of resource management in UEC.

[

158

]

UAV-Enabled Mobile Edge-

Computing for IoT Based on

AI: A Comprehensive Review

2021

This paper presents a review of deep learning and machine learn-

ing techniques used in various applications of UEC.

[48]

UAV-Enhanced Edge Intelligence: A

Survey

2022

The paper introduces and presents various applications of UAV-

enhanced edge intelligence systems. It also discusses some chal-

lenges and future research directions for UAV-enhanced Edge

Intelligent systems.

[47]

Survey on computation ooading in

UAV-Enabled mobile edge computing

2022

The paper presents a comprehensive survey of computational

ooading research in various UEC systems. It compares existing

algorithms qualitatively to assess their features and performances.

[

109

]

A comprehensive survey on aerial

mobile edge computing: Challenges,

state-of-the-art, and future directions

2022 The paper provides an extensive survey of UAV optimization prob-

lems of UAV-enabled mobile edge computing with applications of

Machine Learning techniques.

[

108

]

A survey of mobility-aware Multi-

access Edge Computing: Challenges,

use cases and future directions

2023

This paper comprehensively reviews the service migration, task

ooading, resource allocation and content caching problems in

EC. The paper briey explores the role of UAVs as user equipment.

aspect. A survey on UAV-enhanced edge intelligence is presented in [

] where edge intelligence is

introduced as a key concept in fth-generation networks. The survey presents various applications

of UAV-enabled edge intelligence such as smart agriculture, disaster relief, and intelligent transport

systems. The survey also discusses some key challenges to realising UAV-enabled edge intelligence

such as energy eciency, practical implementation, security, and privacy. A comprehensive review

is carried out in [

158

] which focuses on the role of Machine Learning, and Deep Learning based

methods to enable UEC. The review presents an extensive study of dierent types of UAVs, their

capabilities, and possible applications. Examples of the application areas are agriculture, industry

4.0, environment monitoring, health and emergency, smart cities and smart homes. The review

also presents a classication of the relevant studies in UEC in terms of AI-based approaches. There

are also other surveys in UEC from dierent perspectives, such as security [

], UAVs as user

equipment [

108

], machine learning applications[

109

], etc. Dierent from the existing surveys, our

prime focus is to provide a holistic view of resource management in UEC based on the existing relevant

studies. To the best of our knowledge, there is no existing review that provides a comprehensive

overview of resource management for UEC. A list of the recent surveys is shown in Table 2.

1.4 Literature Methodology

The selection of studies for our research was guided by the principles outlined in the Systematic

Literature Review guidelines [52]. Here, we rst identied and designed the keywords to retrieve

papers from widely-used databases including ACM Digital Library, IEEE Xplore, Wiley, Scopus,

SpringerLink, ScienceDirect and Hindawei: "’Unmanned Aerial Vehicle’ OR UAV, Edge OR Fog,

ooading OR workload OR scheduling OR share OR allocation OR resource OR task OR schedule OR

sharing OR cache OR data OR storage OR provision OR position OR trajectory".

ACM Comput. Surv., Vol. 1, No. 1, Article 1. Publication date: January 2024.

A Survey on UAV-enabled Edge Computing: Resource Management Perspective 1:5

Section 1: Introduction

Section 1.1 Edge Computing and UAV

Section 1.2 Resource Management in UAV

Section 1.3 Related Surveys

Section 1.4 Literature Methodology

Section 1.5 Contributions and Organization

Section 2: Architecture

Section 2.1 Common Architecture

Section 2.2 Collaborations

Section 2.3 Wireless Communication Models

Section 3: Research Directions in UEC

Section 3.1 Resource Provisioning in UEC

Section 3.2 Computation Offloading in UEC

Section 3.3 Resource Allocation in UEC

Section 3.4 Joint Resource Management in UEC

Section 4: Key Techniques & PIs in UEC

Section 4.1 Centralized Methods

Section 4.2 Distributed Methods

Section 4.3 Performance Indicators

Section 5: Challenges and Research Directions

Section 5.1 Architecture

Section 5.2 UAV Trajectory Design

Section 5.3 Real-world Experiments

Section 5.4 Energy Charging and Harvesting

Section 5.5 Security and Privacy

Section 5.6 External Factors

Section 6: Conclusion

Fig. 1. Organization of the Survey

Study selection. We obtained 277 papers using the keywords above and dened inclusion/exclusion

criteria based on [

] to lter out irrelevant or unnecessary studies. Based on these criteria, we

removed 162 papers from the initial set of 277. After reading the full text and applying the criteria,

we obtained 50 papers directly related to resource management in UEC. To further increase study

coverage, we performed backward and forward snowballing on these 50 papers, identifying 7 more

papers. We included a total of 57 studies for our survey. Acknowledging that our selection may

not encompass all relevant literature in the eld, we are condent that our selection encompassed

the majority of essential research studies that reveal the techniques for managing resources in

UAV-enabled edge computing.

Data extraction and synthesis. To ensure a thorough analysis, we conducted a pilot study of

10 papers to gain familiarity with the data to be extracted from the primary studies. Using initial

codes, we iteratively merged them in several rounds to create themes. The analysis was carried out

independently by two authors, with each author analyzing half of the selected papers and then

reviewing the analysis output of the other author. Cases of disagreement were resolved through

discussions among all the authors.

1.5 Contributions and Organization

This survey provides a comprehensive review of the state-of-the-art research on resource manage-

ment in UEC. The key contributions of this work are as follows.

❶

We introduce a three-layered UEC

architecture in section 2, representing a conceptual architecture for managing resources in UEC.

The architecture contains a Things layer, an Edge layer, and a Cloud layer. We then investigate six

types of collaborations that take place in the proposed architecture. The considered collaborations

are a) Things-UAV, b) UAV-Edge, c) Things-Edge, d) Things-UAV-Cloud, e) UAV-Edge-Cloud, and f)

Thigns-UAV-Edge-Cloud. We also discuss the wireless communication models used in UEC.

❷

discover the key research problems of resource management in the context of UEC. In Section 3,

we categorize the research problems into the three following categories: a) computational tasks

and data ooading, b) resource allocation, and c) resource provisioning.

❸

Section 4 identies and

comprehensively reviews the key techniques and performance indicators used for resource manage-

ment in UEC. The key techniques are categorized into two categories: a) centralized methods and

b) decentralized methods. We investigate how these methods are evaluated in the existing work.

In addition, The key performance indicators such as energy consumption, latency, throughput,

ACM Comput. Surv., Vol. 1, No. 1, Article 1. Publication date: January 2024.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

1ASurveyonUAV-enabledEdgeComputing:ResourceManagementPerspectiveXIAOYUXIA,SchoolofComputingTechnologies,RMITUniversity,AustraliaSHEIKMOHAMMADMOSTAKIMFATTAH,CentreforResearchonEngineeringSoftwareTech-nologies,UniversityofAdelaide,AustraliaMUHAMMADALIBABAR,CentreforResearchonEngineeringSoftwareTechnol...

展开>> 收起<<

A Survey on UAV-enabled Edge Computing Resource Management Perspective.pdf

共36页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

A Survey on UAV-enabled Edge Computing Resource Management Perspective

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: