1 s2.0 S0360835221000784 Main

Computers & Industrial Engineering 155 (2021) 107174
Contents lists available at ScienceDirect
Computers & Industrial Engineering

journal homepage: www.elsevier.com/locate/caie
The evolution of the Internet of Things (IoT) over the past 20 years
Jianxin Wang a, Ming K. Lim b, c, *, Chao Wang d, Ming-Lang Tseng e, f, g
a
College of Mechanical Engineering, Chongqing University, Chongqing, China
b
College of Mechanical Engineering, Chongqing University, Chongqing, China
c
Centre for Business in Society, Coventry University, Coventry, UK
d
Research Base of Beijing Modern Manufacturing Development, College of Economics and Management, Beijing University of Technology, Beijing, China
e
Institute of Innovation and Circular Economy, Asia University Taiwan, Taichung, Taiwan
f
Department of Medical Research, China Medical University, Taichung, Taiwan
g
Faculty of Economics and Management, University Kebangsaan Malaysia, Malaysia
A R T I C L E I N F O A B S T R A C T
Keywords: To reveal the origin of the IoT, evaluate its mainstream research topics, and discuss the challenges facing the IoT
Internet of things in the future, this paper conducts a bibliometric study of the IoT from 2000 to 2019. First, this paper presents a
Bibliometric basic bibliometric overview of the IoT. Second, co-citation, coupling and cluster analysis methods are used to
Network analysis
analyse collaboration networks, and VOSviewer is used to visualize the networks. Third, biblioshiny is used to
Thematic trends analysis
analyse the thematic trends of IoT. Finally, this paper discusses IoT challenges and provides some suggestions to
VOSviewer
Biblioshiny address them. The limitations of this paper are also summarized. Research results show that, the mainstream
studies in this field mainly focus on IoT security, wireless sensor networks, IoT management, IoT challenges and
privacy. In addition, the thematic evolution of keywords shows that security and algorithm issues have become
basic themes in the field of IoT research in recent years.
1. Introduction among different people, machines, tasks and knowledge by using the
Internet and sensors (Dang, Piran, Han, Min, & Moon, 2019; Tu, Lim, &
The Internet of Things (IoT) is an extended and expanded system Yang, 2018a). IoT-related research has been extended by some scholars
network based on the Internet, and its ultimate goal is to achieve real- to the Internet of Service (Kwak, Cho, Shin, & Yang, 2020), Internet of
time interaction among things, machines and humans through various Machine (Gazis, 2017), Internet of People (Li, 2017) and Internet of
advanced technological means. The earliest literature on the IoT was Knowledge (Lim, Tseng, Tan, & Bui, 2017; Santoro, Vrontis, Thrassou, &
published in 2002; Schoenberger (2002) first designed the application of Dezi, 2018). With the progress of science and technology, the IoT is
the IoT in stores, and he stated that tiny wireless chips enable stores to expected to achieve large-scale applications in the home and public
have eyes. After nearly 20 years of development, increasing numbers of service markets (Bouzembrak, Klüche, Gavai, & Marvin, 2019). IoT
government officials, corporate executives and researchers tend to applications make important contributions to reducing environmental
believe that the IoT is an important technology for improving our living pollution caused by human activities and increasing the economic
environment and quality of life (Atzori, Iera, & Morabito, 2010; Gubbi, development of countries (Lim, Wang, Wang, & Tseng, 2020; Su, Yang,
Buyya, Marusic, & Palaniswami, 2013; Lim, Xiong, & Lei, 2020; Su, Bai, & Yang, 2018; Tseng, Lim, & Wu, 2019).
Sindakis, Zhang, & Yang, 2020). A market research report showed that To achieve this potential growth, the innovative growth of various
the global IoT market reached $1.90 billion in 2018, and this number is emerging technologies and services needs to keep pace with market
expected to reach $11.03 billion in 2026 (Panetta, 2016). The European demand growth (Li, Lim, Tan, Lee, & Tseng, 2020; Lim & Jones, 2017;
Union (EU), the United States (USA), China and other countries have Zhang, Yang, Zheng, & Su, 2019). Several review articles have covered
also formulated IoT-related action plans. These policies and plans different aspects of the IoT. For example, Atzori et al. (2010) summa
mainly include the IoT-An action plan for Europe and IoT development rized the IoT from the aspects of technology, application scenarios and
plans 2016–2020. major challenges and believed that the multidisciplinary collaboration
The concept of the IoT usually refers to achieving smart connections of telecommunications, informatics, electronics and society promotes
* Corresponding author at: College of Mechanical Engineering, Chongqing University, Chongqing, China.
E-mail addresses: wangjianxin@cqu.edu.cn (J. Wang), ming.lim@cqu.edu.cn (M.K. Lim), Tsengminglang@asia.edu.tw (M.-L. Tseng).
https://doi.org/10.1016/j.cie.2021.107174
Received 3 June 2020; Received in revised form 23 January 2021; Accepted 29 January 2021
Available online 11 February 2021
0360-8352/© 2021 Elsevier Ltd. All rights reserved.
J. Wang et al. Computers & Industrial Engineering 155 (2021) 107174
the IoT to achieve multidisciplinary comprehensive development. Gubbi objectively processing thousands or even tens of thousands of scientific
et al. (2013) reviewed the basic concepts, architectural elements and studies; in particular, visualization functions enable authors to clearly
future development trends of the IoT and proposed a cloud-centric understand the trends of publications (Ghadimi, Wang, & Lim, 2019;
framework for the global IoT. However, the above article did not re Wang et al., 2020). Bibliometric analysis methods have been applied
view the protocols that are used for the IoT. Hence, Al-Fuqaha, Guizani, successfully in different fields, including technology foresight evalua
Mohammadi, Aledhari, and Ayyash (2015) further analysed the tech tion (Gibson, Daim, Garces, & Dabic, 2018), cloud computing research
nologies, protocols and application scenarios of the IoT. All these evaluation (Cai, Lu, Wang, & Xing, 2015), and journal evaluation
research results provide an important reference for researchers and (Wang, Lim, & Lyons, 2019). Hence, three bibliometric methods, co-
programmers to design innovative IoT application scenarios. citation, coupling, and cluster methods, are applied here.
After over 20 years of development, IoT-related research has been
extended to different fields, including smart medical care (Al-Turjman, (a) Co-citation method. Co-citation was put forward by Henry Small,
Nawaz, & Ulusar, 2020; Tu, Lim, & Yang, 2018b), smart agriculture an American intelligence scientist, in 1973 (Small, 1973). The
(Sinha, Shrivastava, & Kumar, 2019), smart supply chains (Manavalan & necessary condition for two articles to form a co-citation rela
Jayakrishna, 2019; Muñuzuri, Onieva, Cortés, & Guadix, 2020), smart tionship is that both articles appear in the references of a third
transportation (Babar & Arif, 2019; Wang, Lim, Zhan, & Wang, 2020) article.
and smart cities (Liu, Bi, & Liu, 2020). The IoT concept has been widely (b) Coupling method. The concept of publication coupling was first
recognized and applied in different fields. However, the IoT-related proposed by Kessler in 1963. Publication coupling refers to the
research results do not explore the inherent development rules and existence of at least two publications that jointly refer to the same
research trends of the IoT. In particular, few studies have been con publication; then, the relationship between two or more publi
ducted to reveal the origin of the IoT, evaluate its mainstream research cations is named coupling.
topics and discuss the challenges that will be faced by the IoT in the (c) Cluster method. Clustering refers to dividing a certain number of
future based on a bibliometric method. In addition, the progress of IoT elements into different groups according to those elements’ de
technology is inseparable from the support of related theories and grees of similarity; the different groups are named clusters (Lee &
methods, and increasingly scholars and practitioners are eager to learn Lee, 2018). After clustering, the elements distributed in the same
more about the development status of the IoT by reading publications. cluster have higher similarity, and the elements distributed in
Hence, it is time for a systematic review and outlook for IoT develop different clusters have lower similarity.
ment over the past 20 years. To achieve this goal, this study mainly
addresses the following four urgent issues: Several tools enable visualization in this study, such as VOSviewer,
which is a professional visualization software developed by Nees and
1) What is the main bibliometric overview of the IoT? Waltman (2020), and biblioshiny, which is a Java software developed by
2) What are the collaboration networks of IoT research? Massimo Aria from the University of Naples Federico (Aria, 2020). In
3) What are the thematic trends of IoT development? this paper, co-citation, coupling, and cluster analysis methods based on
4) What are the main challenges and solutions for the IoT? VOSviewer and biblioshiny are used to analyse IoT-related publications.
To address the above four core concerns of academic researchers, in 2.2. Data sources
this paper, co-citation, coupling and cluster analysis methods are used to
evaluate the development and research trends of the IoT, and VOS Web of Science (WoS), Scopus, and Google Scholar are some of the
viewer and biblioshiny are used to outline a profile of the IoT. The most famous academic resource systems in the word (Hu, Wang, Ni, &
findings of this research help relevant researchers, entrepreneurs and Liu, 2020). Most of the publications included in the Google Scholar
governments have a clearer understanding of the development of the IoT database are automatically retrieved by machines, and this process leads
in the last 20 years and in the future. For researchers, the results are to the loss of key information of some publications. The three main
helpful for understanding the thematic trends and important journals in evaluation indicators used in the Scopus database are Source Normal
the field of the IoT. For entrepreneurs, the results are helpful for finding ized Impact per Paper (SNIP), Impact per Publication (IPP), and SCI
well-known research institutions and developing a competitive IoT mago Journal Rank (SJR), and the two main evaluation indicators used
market after understanding the collaboration networks of the IoT. For in the WoS database are impact factor (IF) and 5-year IF. The IF indicator
governments, the results are helpful for formulating more professional is recognized by an increasing number of scientific research institutions.
action plans based on the development trends of the IoT. The sections of In addition, the WoS database includes SCI & SSCI documents. SCI &
this paper are structured as follows: Section 2 presents the methodology SSCI are citation index publications published and edited by the USA
and data sources. Section 3 shows a basic bibliometric overview of the Institute of Scientific Information (ISI). This retrieval system was foun
IoT. Section 4 shows the collaboration network analysis of the IoT. ded in 1964, and the publications are divided into print, CD and online
Section 5 presents the thematic trends as well as challenges of the IoT, versions, containing more than 8700 core academic journals that have
and finally, Section 6 presents the conclusion. the most influence in various fields, including natural sciences and en
gineering technology. Based on consideration of the advantages of the
2. Methodology and data sources WoS database, this paper uses WoS as the main database for the litera
ture evaluation of IoT publications.
2.1. Methodology As of March 28, 2020, the total number of IoT-related publications
that belong to SCI or SSCI in WoS-Core Collections between 2000 and
Structured literature review methods have been used to analyse and 2019 was 9589, including conference papers (5589, 58.29%), research
review scientific publications. These methods have some advantages, articles (3330, 34.73%), editorial materials (360, 3.75%), review arti
including that authors use them to conduct in-depth analysis for publi cles (193, 2.01%), meeting abstracts (41, 0.43%), news items (26,
cations (Egger & Masood, 2020; Wang, Ghadimi, Lim, & Tseng, 2019). 0.27%), book reviews (24, 0.25%), corrections (16, 0.17%), letters (6,
However, a limitation of these methods is that the time spent by authors 0.06%), retracted publications (3, 0.03%), and a book chapter (1,
increases exponentially as the number of studies increases (Wang, Zhao, 0.01%). Considering the completeness of the research content, methods,
Vilela, & Lim, 2019). In addition, it is difficult to eliminate the influence and results of different types of publications, research articles and re
of author subjective factors on the analysis results (Addo-Tenkorang & view articles (3523) are analysed. These publications were drawn from a
Helo, 2016). In contrast, bibliometric analysis methods enable total of 10,210 authors, 2851 institutions, and 94 countries. Notably,
2
Table 1
Top 20 journals by the number of publications.
R Journal N IF- IF- IF- IF-5Y FY TC TC/N TC/Y >500 [300, [100, < Y
2010 2015 2018 500] 300] 100
1 IEEE Access 375 – 1.270 4.098 4.540 2014 20,879 55.68 3479.83 2 0 6 367 2013
2 IEEE Internet of Things J. 363 – – 9.515 11.216 2014 6,119 16.86 1019.83 2 2 7 352 2014
3 Sensors 217 1.452 1.570 2,295 3.302 2012 46,222 213.00 5777.75 0 0 1 216 2001
4 Future Generation Computer 151 2.371 2.430 5.768 5.670 2013 10,230 67.75 1461.43 1 0 0 150 1995
Systems-the Int. J. of eScience
5 Int. J. of Distributed Sensor 142 0.067 0.906 1.614 1.461 2012 4,131 29.09 516.38 0 0 1 141 2004
Networks
6 IEEE Communications 88 2.837 5.125 10.356 12.091 2011 24,753 281.28 2750.33 0 0 10 78 1995
Magazine
7 IEEE Transactions on 68 1.627 4.708 7.377 8.423 2013 13,187 193.93 1883.86 1 1 5 61 2004
Industrial Informatics
8 Wireless Personal 62 0.507 0.701 0.929 0.959 2009 5,256 84.77 477.82 0 1 2 59 2000
Communications
9 J. of Network and Computer 53 0.660 2.331 5.273 4.744 2012 6,959 131.30 869.88 0 1 5 47 1995
Applications
10 Computer Networks 49 1.176 1.446 3.030 2.989 2010 10,122 206.57 1012.20 1 3 2 43 1997
11 Wireless Communications 47 0.810 0.922 1.396 1.364 2016 3,421 72.79 855.25 0 0 0 47 2002
Mobile Computing
12 Computer 44 1.812 1.115 3.564 2.833 2011 7,260 165.00 806.67 0 0 3 41 1995
13 Int. J. of Advanced Computer 44 – – – – 2015 – – – 0 0 0 44 –
Science and Applications
14 IEEE Consumer Electronics 41 – – 3.273 2.446 2012 636 15.51 79.50 0 0 1 40 2014
Magazine
15 Computers Electrical 39 0.484 1.084 2.189 2.337 2011 3,855 98.85 428.33 0 0 1 38 1995
Engineering
16 Cluster Comp. the J. of 37 0.679 1.514 1.851 1.359 2017 2,099 56.73 699.67 0 0 0 37 2005
Networks Software Tools and
App
17 Personal and Ubiquitous 37 1.137 1.498 1.735 2.061 2011 2,371 64.08 263.44 0 0 2 35 2006
Computing
18 Ad Hoc Networks 36 1.592 1.660 3.490 3.336 2012 5,084 141.22 635.50 1 0 4 31 2007
19 IEEE Network 34 1.934 2.899 7.503 7.344 2011 4,228 124.35 469.78 0 0 3 31 1995
20 Multimedia Tools and 34 0.914 1.331 2.101 1.876 2015 8,199 241.15 1639.80 0 0 0 34 1996
Applications
Note: R represents the quantity ranking. N represents the number of IoT articles indexed in WoS. IF represents the impact factor. FY represents the year when IoT
articles were first indexed. TC represents total citations. > and [,] represent intervals of the number of citations of IoT articles, counted on March 18, 2020. Y represents
the earliest year that is reported by InCites Journal Citation Reports (JCR).
institutions from Hong Kong and mainland China are combined into one (a) The journal with the highest 5-year IF is IEEE Communications
group named China, and institutions from Northern Ireland, Wales, Magazine with 12.091, followed by IEEE Internet of Things Journal
Scotland and England are combined into one group named the UK. with 11.216. The journal with the largest number of IoT articles is
IEEE Access (375, 8.05%), followed by IEEE Internet of Things
3. Bibliometric overview Journal (363, 7.79%). The journal with the most total citations is
Sensors, with 46,222 citations. IEEE Communications Magazine
3.1. Journal overview tends to accept papers on hot topics.
(b) IEEE Internet of Things Journal is the youngest journal and was
A total of 3523 IoT-related publications published in 2000–2019 are launched in 2014. This journal mainly publishes IoT-related ar
analysed. In terms of the number of IoT publications, the top 20 journals ticles on topics including communication protocol design and
are shown in Table 1. The IFs in Table 1 are from the annual Journal optimization, innovative practices and application, major tech
Citation Reports (JCR). The following two features are obtained from nological innovation, system framework and architecture of the
Table 1: IoT, social management system innovation, etc.
1600
1436 120
104
1400
1177 100
1200
1000
80
809 60
800 60
591 42
600 40 32
400 338
239 20 12
187 4 6
200
0
0
2013 2014 2015 2016 2017 2018 2019 2013 2014 2015 2016 2017 2018 2019
a) Research articles b) Review articles

Fig. 1. The change trends of the numbers of different types of publications.
3
Fig. 2. Analysis of the reasons for the changes in the number of IoT articles.
Table 2
Citation structure in terms of IoT-related papers according to WoS.
Year ≥300 ≥200 ≥100 ≥50 ≥20 ≥10 N TC TC/N h-index CA
Total 31 49 118 276 670 1083 3523 74,720 21.21 109 37,095
2019 0 0 0 4 28 67 1264 2958 2.34 23 2393
2018 0 0 8 24 118 224 887 8698 9.81 42 6423
2017 1 2 11 45 131 239 506 9303 18.39 48 6999
2016 3 7 19 53 136 199 342 10,748 31.43 51 8376
2015 6 7 17 48 81 123 185 9755 52.73 48 7662
2014 10 13 23 41 78 101 141 12,032 85.33 44 8904
2013 4 5 11 24 49 66 97 7869 81.22 35 6629
2012 2 4 7 12 16 23 44 3220 73.18 17 2864
2011 1 5 13 13 17 22 35 2796 79.89 18 2336
2010 2 4 6 8 10 11 13 6481 498.54 11 6037
2009 1 1 2 3 5 6 6 569 94.83 6 545
2008 0 0 0 0 0 0 0 0 0 0 0
2007 0 0 0 0 0 0 0 0 0 0 0
2006 0 0 0 0 0 0 0 0 0 0 0
2005 0 0 0 0 0 0 0 0 0 0 0
2004 1 1 1 1 1 2 2 285 142.5 2 285
2003 0 0 0 0 0 0 0 0 0 0 0
2002 0 0 0 0 0 0 1 6 6 1 6
2001 0 0 0 0 0 0 0 0 0 0 0
2000 0 0 0 0 0 0 0 0 0 0 0
Note: N represents the number of IoT articles indexed in WoS. TC represents total citations. h-index represents that at most, h publications have been cited at least h
times. CA represents the article citations; all of these data were counted on March 28, 2020.
3.2. Quantity distribution published the “Twelfth Five-Year” Development Plan Report on the IoT
in 2010 (Most, 2012); this report provided a detailed plan for the key
The number of publications over the years indicates research results development directions of the IoT from 2011 to 2015. The third stage
and trends in the IoT field. As of January 16, 2020, this paper classifies was (2015, 2019], when 2999 publications were published in WoS, and
3523 publications from the WoS database; the published numbers of the annual increase in publications reached more than 85.13% in this
research and review articles from 2013 to 2019 are shown in Fig. 1. stage.
The first article in the IoT field appeared in 2002, when Schoen
berger (2002) mainly introduced application scenarios of the IoT in 3.3. Citation structure
future smart supermarkets, including how to use sensors for detection in
the operation of a supermarket. After 2009, these publications showed a Table 2 shows the citation structure in terms of IoT-related publi
growing trend. In contrast, review articles appeared relatively late, and cations, where the column of “≥300” represents the number of articles
the first review article appeared in 2012. The change trend of the that have been cited more than 300 times. The statistical results show
number of publications is analysed and shown in Fig. 2. that these IoT-related publications have received 74,720 citations in the
Fig. 2 shows that the development of the IoT has gone through three last 20 years, and the most-cited publications were published in 2014,
stages. The first stage was (2002, 2009], when nine publications were with a total of 12,032 citations, accounting for 16.10%. The total
published. The International Telecommunication Union (ITU) released a number of citations showed a trend of first increasing and then
report on the IoT in 2005 (ITU, 2003); after that, the IoT gradually decreasing around 2014. This phenomenon shows that academic
entered a slow development period. The second stage was (2009, 2015], research on the IoT made an influential breakthrough in 2014. Repre
and many countries issued action plans on the IoT during this stage. For sentative study results during this period are about the application of the
example, the EU released the “Internet of Things - An action plan for IoT in smart cities and industries (Zanella, Bui, Castellani, Vangelista, &
Europe” in 2009 (Brussels, 2009). The Chinese government also Zorzi, 2014). This is followed by 2016, with a total of 10,748 citations,
4
Table 3 Telecommunications, with 97 publications. It is worth mentioning that

The statistical analysis of institutions. the CAS is the highest academic institution in China (Zhang & Zhao,
Ranking Institution Article Proportion 2019), is the highest scientific and technological institution and is a
quantity natural science and high-tech comprehensive research centre. At pre
1 Chinese Academy of Sciences 109 3.09% sent, the CAS mainly focuses on the research of advanced sensors and
2 Beijing University of Posts 97 2.75% modules, communication systems and devices, optical communication
Telecommunications devices and modules, wireless information systems and networks.
3 King Saud University 72 2.04%
4 University of Electronic Science 61 1.73%
Technology of China 4. Network analysis
5 Dalian University of Technology 47 1.33%
6 Vellore Institute of Technology 46 1.31% Network analysis is an important part of bibliometric analysis and
7 Huazhong University of Science 43 1.22%
mainly reflects the collaborative relationships among IoT publications.
Technology
8 University of California System 39 1.11%
Next, analyses using co-authorship, citation and bibliographic coupling
9 Shanghai Jiao Tong university 38 1.08% networks are presented.
10 University of Science and Technology 38 1.08%
Beijing
4.1. Co-authorship network analysis
accounting for 14.38%, but these data may change over time, as pub Fig. 3 shows the collaboration network of different authors in terms
lications still take time to reach their highest citation levels (Wang et al., of publications. The nodes represent author names, the links represent
2020). In addition, the average number of citations per year is 21.21 the co-authorship relationships between different authors, and the node
from 2002 to 2019. The most-cited publication was published in 2010, sizes represent each author’s number of publications. The co-authorship
with an average number of citations of 498.54. Moreover, 0.88% of network analysis results show that Guizani, Mohsen is the most influ
these publications have obtained more than 300 citations in WoS, 1.39% ential author in terms of the total link strength. Guizani, Mohsen mainly
of these publications have obtained over 200 citations, 3.35% of these studies IoT technologies innovative and practices (Al-Fuqaha et al.,
publications have obtained over 100 citations, and 7.84% of these 2015). Atzori, Luigi is the most influential author, with 6529 citations.
publications have obtained over 50 citations. Note that the single most- The data and network structure (Fig A1) show that collaborative
cited paper was published in 2010, with 5153 citations. This was a re relations between developing countries and other countries are
view article that pointed out that the IoT comprises multiple technolo becoming more frequent, which is one of the causes of the rapid
gies and gave a detailed introduction to the challenges faced in the development in developing countries. In contrast, institutions in devel
development of the IoT (Atzori et al., 2010). oped economies rarely take the initiative to cooperate with institutions
in other economies; they tend to choose institutions that are better than
3.4. Institutional structure themselves as partners. The country co-authorship network is shown in
Fig A2. The results show that the UK, Saudi Arabia, the USA, France and
In terms of the quantity of journal articles published, the Chinese Iran are becoming research centres for the IoT. In addition, these
Academy of Sciences (CAS) has the largest number of IoT articles. The countries include developed economies and developing economies,
quantity statistics of institutions are shown in Table 3. Table 3 shows which indicates that there is no theoretical leading region for the study
that the CAS has the most published journal articles, with 109 articles, of the IoT. The country collaboration map is shown in Fig. 4.
23 h-index articles, and 4198 citations, and the average annual number The collaboration map (Fig. 4) shows that collaborative research
of citations is 38.43. This is followed by Beijing University of Posts and between different countries has become a mainstream trend. The
Fig. 3. The author co-authorship network. Note: The network is visualized by VOSviewer. The whole network consists of 156 nodes, 17 clusters and 344 links. The
total link strength value is 659.
5
Fig. 4. Collaboration map of the top five countries. Note: The network is visualized by biblioshiny. The red links represent the collaborative relationships between
different countries. The blue colour represents the number of publications, where a darker colour indicates a greater number of publications.
Fig. 5. The source citation network. Note:

The network is visualized by VOSviewer. The
whole network consists of 116 nodes and
1789 links, and the total link strength of this
network is 8035. Those journals marked in
red tend to accept publications with general
topics in the IoT. The nodes marked in blue
pay more attention to the related topics of
the integration of the IoT and business. The
nodes marked with green tend to pay more
attention to the application of the IoT in
green and clean products. Nodes marked in
yellow tend to focus on sensor topics.
collaborative relationship between the USA and China is the most greater results than individual research.
frequent, with 204 collaborations. This is followed by China and the UK,
with 78 collaborations. In addition, the country collaboration map
shows that China, the USA, the UK, Australia, Canada and Korea have 4.2. Citation network analysis
extensive collaborative relations with other countries in the world.
Notably, China, the USA, the UK, Australia and Canada have published In this subsection, the source citation network is constructed, as
2544 IoT-related papers, accounting for 72.21%. This also shows that shown in Fig. 5. This network shows that the top ten journals by the
cooperative research between countries is more easily able to achieve number of links are Computer Networks (106), IEEE Internet of Things
Journal (105), Future Generation Computer Systems-the International
6
Table 4
Detailed information for the authors citations.
R Author W1 W2 W3 W4 W5 S1 S2 S3
1 Atzori, Luigi 80 403 14 6529 34.13 2014 466.36 2.44

2 Iera, Antonio 79 365 10 6241 31.37 2014 624.10 3.14
3 Morabito, Giacomo 78 343 8 6135 26.82 2014 766.88 3.35
4 Buyya, Rajkumar 66 173 10 3933 61.87 2017 393.30 6.19
5 Xu, Lida 65 241 14 3617 66.41 2014 258.36 4.74
6 Li, Shancang 57 178 7 2410 44.25 2014 344.29 6.32
7 Guizani, Mohsen 71 302 22 2131 74.98 2018 96.86 3.41
8 Vasilakos, Athanasios v. 66 204 17 2084 59.15 2015 122.59 3.48
9 Perera, Charith 51 179 12 1861 29.11 2016 155.08 2.43
10 Georgakopoulos, Dimitrios 48 138 9 1589 23.17 2015 176.56 2.57
Note: W1 represents the weight <Links>; W2 represents the weight <Total link strength>; W3 represents the weight <Publications>; W4 represents the weight
<Citations>; W5 represents the weight <Norm. citations>; S1 represents the score <Avg. pub. year>; S2 represents the score <Avg. citations>; S3 represents the score
<Avg. norm. citations>.
Fig. 6. The citation analysis results based on research institution. Note: The network is visualized by VOSviewer. The maximum number of institutions per pub
lication was set to 25, and 22 out of the 2851 institutions met the thresholds.
Journal of eScience (97), IEEE Access (97), IEEE Communications Surveys and trustworthiness management (Nitti, Girau, & Atzori, 2014). The
and Tutorials (93), Sensors (92), Ad Hoc Networks (89), Journal of Network second-largest link is Iera, Antonio from the University of Reggio Cala
and Computer Applications (89), IEEE Transactions on Industrial Infor bria. His team shares the same research themes as the team of Atzori,
matics (83), and IEEE Communications Magazine (80). These journals are Luigi, including the SIoT (Atzori, Iera, & Morabito, 2014; Atzori et al.,
the core journals that accept IoT-related papers. 2012). The detailed information of author citations is shown in Table 4.
From the source citation network results, some valuable conclusions In terms of citations, the influential authors also often have collab
can be obtained, including that these important journals maintain strong orative relationships, such as Atzori, Luigi. Iera, Antonio and Morabito,
citation relationships with each other. The nodes representing these Giacomo have a partnership, and the most impactful article they ever
journals are mostly red, which indicates that the types of articles published was “The Internet of Things: A survey”, which was published
accepted by these journals have a certain similarity and that each in Computer Networks (Atzori et al., 2010). This article in the WoS
journal has not formed a distinctive feature theme. The key indicator for database has been cited 5153 times as of March 28, 2020. The citation
judging whether different journals are similar is whether there is a analysis results are shown in Fig. 6.
relationship between citation intensity and citation direction among the Fig. 6 shows the related collaborative relationships between different
publications they accept. This phenomenon may change with the pas research institutions. The clustering results show that the CAS and King
sage of time, as different journals may tend to accept different types of Saud Univ have extensive collaborative relationships with research
research articles, including technology, applications and algorithms. structures in other parts of the world. The CAS and King Saud Univ are
Detailed information on these journals can be found in Table 1. located in developing economies, and these results show that developing
The author citation network is shown in Fig A3. The network analysis economic regions play a key role in promoting collaborative research on
results (Fig A3) showed that the largest link is Atzori, Luigi from the the IoT. The country citation network is shown in Fig. 7.
University of Catania. His team mainly focuses on the social IoT (SIoT) Fig. 7 shows that the top ten countries by normalized number of
(Atzori, Iera, & Morabito, 2011; Atzori, Iera, Morabito, & Nitti, 2012) citations are China (1311), the USA (999), the UK (335), Australia (307),
7
Fig. 7. The country citation network. Note: The network is visualized by VOSviewer. The whole network consists of 70 nodes, 4 clusters and 1612 links, and the total
link strength of this network is 34,994.
Fig. 8. The results of the coupling analysis based on authors. Note: The network is visualized by VOSviewer. The nodes represent publications; the edges represent
citation relationships between publications. The number of edges connected to a publication determines the size of the node. The larger the node is, the stronger the
citation relationship between the publication and other publications.
8
Table 5 4.3. Bibliographic coupling network analysis

The most weighted literature data in these five categories based on VOSviewer.
Label X Y Cluster T1 T2 T3 T4 The coupling analysis results (Fig A4) based on sources show that the
largest weight is obtained for IEEE Internet of Things Journal, followed by
Atzori 0.3386 0.3515 2 676 11,810 731 2.32%
et al. IEEE Access, Future Generation Computer Systems-the International Journal
(2010) of eScience, IEEE Transactions on Industrial Informatics, and Sensors. Their
Zhang and − 0.1043 − 0.2659 1 617 3379 216 0.66% weights are 561.41, 334.75, 329.86, 142.31, and 119.40, and their
Wen shares of the total value reach 20.72%, 12.36%, 12.18%, 5.25%, and
(2017)
Roman, − 0.2889 0.6223 3 601 4703 195 0.92%
4.41%, respectively. It should be noted that IEEE Communications Sur
Zhou, veys and Tutorials is an online journal published by the IEEE Commu
and nications Association; its impact factor has reached 25.222 in the past
Lopez five years, and its content covers all aspects of the communications field.
(2013)
The IEEE Internet of Things Journal and IEEE Access have many re
Li et al. − 0.593 0.3579 5 574 3326 166 0.65%
(2011) lationships with other journals, and these results show that they are
Lee and 0.8052 0.335 4 462 1724 102 0.34% playing an important role in influencing the future research trends of the
Lee IoT. The coupling of publications reflects the relationship between two
(2015) cited publications. The author bibliographic coupling network is shown
T1: weight <Links>; T2: weight <Total link strength>; T3: weight <Citations>; in Fig. 8.
T4: percentage. Fig. 8 shows the coupling analysis results. These results show that the
most influential article in the field of the IoT is Atzori et al. (2010),
India (293), Italy (257), South Korea (245), Canada (184), Saudi Arabia which introduced the key technologies for implementing the IoT and the
(168), and France (153). Developed economies accounted for 70% of the development direction of the IoT, and that article provides a good
top ten, and the sum total of IoT publications in developed economies reference for the research of the IoT. The weight value of that article in
was 2480, accounting for 58%. This shows that developed economies the network is 11,810, and the proportion is 2.32%. This is followed by
play an important role in scientific research in the field of the IoT. In Gubbi et al. (2013), which was published in Future Generation Computer
addition, developing economic regions are playing a key role in pro Systems; that research mainly focuses on the constituent elements of the
moting collaborative study in the IoT field. IoT and future development directions. Gubbi et al. (2013) has a weight
of 6787 in the network, and the ratio is 1.33%. After co-citation analysis,
these studies are divided into five categories, and the literature data with
the highest weights in these five categories are shown in Table 5.
Fig. 9. The results of coupling analysis based on countries. Note: The network is visualized by VOSviewer. The nodes represent countries; a larger node indicates that
the node has greater influence on other nodes. The lines represent the mutual relationships between those nodes; the different colours of the lines represent
different clusters.
9
Fig. 10. Keyword co-occurrence network. Note: The network is visualized by VOSviewer. The minimum number of occurrences of a keyword is set to 6.
The number of studies published by scholars is one of the important further analyse the mutual coupling relationships between those nodes
indicators, that enables us to reflect whether scholar belong as experts in and the characteristics, this sub-section selects several representative
the IoT field. It has a positive effect for readers to quickly learn about countries to further analyse the coupling of the literature. The publica
experts in the IoT field by analysing the number of articles published. tion coupling network is shown in Fig A5.
Table 5 shows that these authors have a considerable number of publi The following two features can be obtained from Fig A5: First,
cations. As far as the number of articles cited is concerned, most of are different countries tend to have coupling relationships with other
from Atzori et al. (2010) who comes from the University Cagliari; he has countries that have a greater influence on the IoT, which has led to the
published 46 journal articles related to the IoT. In-depth mining of the emergence of widespread collaboration between countries that have
author’s relevant information, in this study showed that most of these greater influence and other countries that have less influence in the IoT
authors are university professors. In addition, most authors have field (Fig A5. a and b confirm this conclusion). Those countries with
collaborative relationships and come from the same institution, for greater influence include the USA, China, the UK, and India. Second,
example, Atzori et al. (2010) published an article entitled “The Internet countries with moderate influence tend to have coupling relationships
of Things: A survey”. with those countries that have a greater influence rather than with
In-depth collaboration between scientists and researchers from countries with less influence. As a result, countries that have moderate
different countries in the IoT field has become a mainstream trend, and influence have more citation collaboration with countries that have a
this in-depth collaboration phenomenon has a positive effect on pro high relative influence (Fig A5. c and d confirm this conclusion). These
moting the rapid rise of IoT applications in different scenarios. The regions include Japan and Singapore.
coupling analysis results based on countries are shown in Fig. 9.
Fig. 9 shows the results of the literature coupling. According to the 5. Thematic trend and challenge analysis results
weight ranking, the top five most influential countries are China, the
USA, the UK, South Korea and India. The total link strength of these A total of 11,899 keywords in 3523 publications from 2000 to 2019
countries with other countries has reached 499,486; 409,396; 224,934; are analysed. A total of 2531 keywords appeared twice, 1401 keywords
202,134; and 179,176, respectively. It is important to note that the total appeared three times, and 962 keywords appeared four times. The most
connectivity is 3,432,852. Their proportions are 14.55%, 11.93%, frequently used keyword is “internet of things”, used 1710 times, fol
6.55%, 5.89%, and 5.22%, respectively. The coupling results show that lowed by “iot” 674 times, “security” 456 times, “internet of things (iot)”
China, the USA, South Korea and India cite large numbers of the same 557 times and “wireless sensor networks” 324 times. Next, the thematic
literature, which indicates that the development of the IoT in these trends and challenges are analysed.
countries has the same literature reference foundation. Aiming to
10
Table 6 5.1. Keyword co-occurrence network analysis

The main keywords identified based on a cluster method.
Cluster Cluster keywords (occurrences ≥ 20) and the most frequent keywords Based on the above keyword data, a cluster method is used to analyse
the co-occurrence features of different keywords in the related IoT
Cluster 35 keywords: internet of things, cloud computing, sensors, big data,
1 sensor, rfid, machine learning, smart city, wireless sensor network, literature in this sub-section, which aims to show the main trends of
middleware, smart cities, web, classification, ontology, interoperability, keyword co-occurrence. The keyword co-occurrence network analysis
deep learning, artificial intelligence, internet, zigbee, smart objects, results are shown in Fig. 10. All keywords are divided into four main
health, identification, prediction, anomaly detection, big data analytics, clusters. Based on these analysis results, keywords with more than 20
data mining, semantic web, data analytics, monitoring, quality of service,
care, recognition, environment, network security, social IoT.
occurrences in each cluster are selected, and the detailed information of
The most frequent keyword: internet of things, occurrences = 1710. each cluster is shown in Table 6.
Cluster 42 keywords: wireless sensor networks, networks, design, systems, Cluster 1 mainly focuses on the IoT, big data, sensor research, and
2 algorithm, optimization, sensor networks, communication, wireless, cloud computing. The most frequent keyword is “internet of things”,
energy, network, energy efficiency, 5 g, energy harvesting, access,
with 1710 occurrences. Cluster 2 mainly focuses on algorithm and
resource allocation, selection, algorithms, networking, internet-of-things
(iot), routing protocol, localization, transmission, discovery, sensor optimization research, including algorithm design for the IoT and
network, wireless networks, allocation, qos, resource allocation, routing protocols. Cluster 3 mainly focuses on IoT management issues
cognitive radio, coverage, game theory, reliability, mobility, related to performance management and supply chain management.
performance analysis, power, clustering, physical layer security, strategy, Cluster 4 pays more attention to security and challenge issues, such as
tracking, machine, routing.
The most frequent keyword: wireless sensor networks, occurrences =
network attacks and privacy. The results are shown in Table 6.
557. To analyse the evolution of keywords in the last 20 years, the
Cluster 34 keywords: management, system, model, framework, technology, keyword co-occurrences with a timeline are shown in Fig. 11. The red
3 performance, information, industrial IoT, technologies, service, nodes represent the most recent emerging topics, which include block
integration, smart, services, future, platform, cyber-physical systems
chain, wireless power transfer (WPT), fog, and intrusion detection. As an
(CPS), innovation, industry 4.0, healthcare, implementation, mobile,
smart grid, smart home, impact, vision, quality, supply chain, context, emerging topic in the field of the IoT, a block chain is a decentralized
opportunities, adoption, analytics, knowledge, standards. core system that has been used in digital cryptocurrencies (Carreño
The most frequent keyword: management, occurrences = 230. et al., 2019; Li, Cai, Deng, Yao, & Wang, 2019). In addition, using radio-
Cluster 40 keywords: security, challenges, privacy, architecture, scheme, frequency WPT to achieve three-dimensional positioning of IoT devices
4 protocol, cloud, authentication, blockchain, trust, fog computing, edge
has also become an emerging topic in research this year. Existing
computing, efficient, industrial internet of things (iiot), secure, attacks,
intrusion detection, protocols, trust management, wsn, coap, health-care, research results show that as long as a smart device is located in the
devices, access-control, issues, rpl, encryption, 6lowpan, cryptography, radiating near-field area, the distance from the anchor point to the target
access control, lightweight, iot applications, mutual authentication, key can be estimated easily (Aziz et al., 2019).
agreement, low-power, cybersecurity, key agreement scheme, survey, of-
the-art, user authentication.
The most frequent keyword: security, occurrences = 456. 5.2. Keyword evolution trends
To reveal the evolution of keyword trends, strategic diagrams are

constructed, as shown in Fig. 12. The strategic diagrams include four
quadrants: the first quadrant represents the thematic keywords
Fig. 11. The keyword co-occurrences with a timeline. Note: The network is visualized by VOSviewer. This network includes 268 nodes and 8741 links; note that the
different colours of the nodes represent different years.
11
Fig. 12. Keyword thematic map for IoT-related articles from 2000 to 2020. Note: The network is visualized by biblioshiny.
belonging to motor themes; the second quadrant represents the thematic 5.3. Challenges and discussion
keywords belonging to highly developed themes; the third quadrant
represents the thematic keywords belonging to emerging themes; and According to the keyword co-occurrence network and keyword
the fourth quadrant represents the thematic keywords belonging to basic thematic evolution results, IoT trust management, security and archi
and transversal themes. tecture have become challenging mainstream research topics. The se
In Fig. 12 (a)–(c), the motor area represents themes that have been curity topic appears 456 times, architecture and framework topics
well developed in the IoT field. The highly developed and isolated area appear 380 times, and trust and privacy topics appear 294 times. Hence,
represents themes that have good internal development. The emerging in this subsection, the challenges of IoT trust management, security and
or declining area represents themes that have weakly developed. The architecture are analysed and discussed in detail, and these challenges
basic and transversal areas represent themes that have weak internal aim to transform the IoT from a hot concept into a good engineered
development. Next, the thematic evolution is analysed based on key viable technological paradigm.
words, and a Sankey diagram based on keyword thematic evolution is a) Trust challenges. There are many nodes involved in an system
shown in Fig. 13. composed of various intelligent devices and transmission protocols
According to the Sankey diagram (Fig. 13) and keyword thematic (Ding, Hu, Ke, Wang, & Chang, 2019; Zhang, Yang, Wang, Li, & Su,
map (Fig. 12) analysis results, the following two features are valuable: a) 2018; Zhang, Yang, Su, & Zheng, 2019). When a single node in the IoT
The basic and transversal themes show that security and algorithm is network is attacked, it is easy to cause paralysis of the IoT system. One of
sues have become basic themes in the IoT field in recent years. From the better ways to solve the collapse issue is to improve the robust
2000 to 2013, the basic themes focused more on networks, design and performance and reliability of the IoT network, so the trust management
wireless sensor networks. From 2014 to 2016, the basic themes focused (TM) issue in the IoT field has attracted the attention of many scholars
more on IoT security and algorithms. Security and privacy challenges for (Ben, Olivereau, Zeghlache, & Laurent, 2013). Fig. 14 shows a schematic
the IoT and potential solutions are introduced in detail in the challenges diagram of the lack of IoT functions caused by some node failures.
and discussion section. b) The highly developed and isolated themes Fig. 14 shows an IoT system consisting of people, vehicles, ware
show that IoT networks and design have developed rapidly in recent houses and computers. Previous research results show that TM plays a
years. From 2000 to 2016, IoT routing protocols and systems developed positive role in building an IoT application service system based on the
rapidly. Rank and Sybil attacks are the mainstream attack methods in effective fusion of multisource isomerism data, qualified service quality,
the IoT. In recent years, research has aimed to provide more secure and strong user privacy protection (Yan, Zhang, & Vasilakos, 2014). To
application scenarios for the IoT, and the design of a more secure and solve the issues of TM in IoT management, Ning, Liu, and Yang (2013)
trustworthy routing protocol has become a main study hotspot and ur provided an IoT system structure solution from four perspectives,
gent issue (Airehrour, Gutierrez, & Ray, 2019). including data, system, network and application security. However,
those specific issues, including HCTI, SSR, DFMT, and TDR, have not
12
Fig. 13. Sankey diagram based on keyword thematic evolution from 2002 to 2020. Note: The network is visualized by biblioshiny. The number of keywords is set to
500, the minimum cluster frequency is set to 3, and the inclusion index is weighted by word occurrences.
Fig. 14. A schematic diagram of the lack of IoT functions caused by some node failures.
been taken into account. Compared with previous research, Sun et al. the security of the IoT, and they also provide common IoT attack
(2018) seemed to use a more micro-level approach in solving security methods and corresponding solutions (Lin et al., 2017). Harbi, Aliouat,
management issues, and Sun et al. (2018) briefly reviewed TM in the IoT Harous, Bentaleb, and Refoufi (2019) reported that because an IoT
from the aspects of authenticity and integrity of sensor data, lightweight network enables access for a large number of devices, this also increases
encryption algorithms and protocols, and key protocols of the physical the risk of an attack on the network. These risks usually cause data loss
layer. However, as Li and Zhou (2011) described, the IoT architecture, in the IoT system; in a bad situation, attacks cause specific functions of
the security of information processing and the protective management the IoT to be lost. Table 7 shows the names of security risk attacks
of personal privacy are unavoidable issues in the building process. Yan mentioned in the current literature, the potential consequences of those
and Wang (2010) studied IoT TM from three aspects, the construction of attacks, and the corresponding solutions.
heterogeneous network models, trust routing and TM design, and pro Table 7 shows the names of security risk attacks, the potential con
vided more detailed guidance for the design of a secure IoT. sequences of those attacks, and the corresponding solutions. The com
b) Security challenges. The IoT also faces security challenges in its mon feature of these attacks is to destroy or maliciously access the IoT
development. Previous investigations and research have fully described system, thereby achieving the purpose of damaging the IoT system
13
Table 7 Table 7 (continued )

The names of security risk attacks, potential consequences and corresponding No. Name Basic description Potential Corresponding
solutions. consequences solutions
No. Name Basic description Potential Corresponding Repetitive Gaining the trust Undermine Set time and
consequences solutions attack of the IoT through the validity of repeat threshold
1 Sleep Keep the node or Some nodes Increase node multiple the certificate
deprivation device awake for a fail and the energy storage malicious inputs
attack long time until it network is capacity and 18 Malicious virus Attacks system by System crash Deploy a reliable
is forced to shut down adopt energy attack disguising itself as firewall
down storage a self-propagating
technology virus
2 Sybil attack Malicious device Data leak, Increase security
has multiple legal certain identification
identities in the functions fail mechanism
system in IoT Table 8
3 Phishing attack Pretend to be a User account Be alert when Structure statistics of the IoT.
phishing website password users go online The structure of the IoT Representative literature
to trick user leaked
information Sensing, accessing, networking, middleware ITU (2003)
4 False data Transfer fake data Executes Detecting false or and application layers
injection attack into IoT wrong error data Perception, network and service layers (or Domingo (2012)
applications feedback application layer)
command Application, network and sensing layers Atzori et al. (2010)
5 Malicious Pretend to steal Confidential Add dynamic Sensing, networking, service and interface Xu, He, and Li (2014)
script attack user confidential data leak monitoring layers
data in mechanism Sensing, networking and application layers Atzori et al. (2010)
installation Sensing, networking, service and interface Li, Xu, and Zhao (2015)
software layers
6 Sinkhole attack Infected device or Damage to Add multiple Topology, architecture, and platform layers Islam, Kwak, Kabir, Hossain, and
node as a circular transmitted security protocols Kwak (2015)
forwarding node data, data Application, transportation and perception Jing, Vasilakos, Wan, Lu, and Qiu
leakage layers (2014)
7 Man-in-the- Maliciously steal Loss of data Deploying a Smart objects/smart devices, hubs, cloud, third Stojkoska and Trivodaliev (2017)
middle attack and control integrity and secure party
communication accuracy communication Application, network, and device layers Gazis (2017)
information protocol Perception, network and application layers Lin et al. (2017)
between two Application, transportation and perception Shin (2014)
normal devices layers
8 Deception Pretend to be a Specific Add multi-factor Application, transport, network, MAC and Jin, Gubbi, Marusic, and
attack normal node or functions fail authentication physical layers Palaniswami (2014)
device to scam in the IoT Perception, network, middleware and Fang et al. (2014)
access system application layers
9 Worm attack Send malicious Cause Modify routing Devices, communication gateways, data centres Hossain and Muhammad (2016)
packets through network protocol and application layer
two malicious congestion Integrated services, applications and sensing Tsai, Lai, Chiang, and Yang
nodes or devices and failure entities layer (2014)
10 Cryptanalysis Maliciously Data leaks in Set a secure Open flow access switch, data plane, internet Sun and Ansari (2016)
attack inferring systems, key encryption and cloud layer
encryption keys functions fail algorithm Smart device, edge/fog and cloud layers Rahmani et al. (2018)
for IoT systems Collection station, data centre and observation Abawajy and Hassan (2017)
11 DDoS attack Bombs network Network Increase network station layer
with very large paralysis protection system Wireless sensor networks, cloud computing and Gubbi et al. (2013)
traffic, occupying applications layer
available
resources
12 Unauthorized Obtain Information Add RFID (Husain & Mohamed, 2019). Sleep deprivation attacks, phishing attacks
attack information by leakage authorized access and false data injection attacks have caused great challenges in the se
accessing RFID mechanism
curity of the IoT. Next, this subsection tries to give some effective so
without
authentication lutions to these attacks. Sleep deprivation attacks cause the edge node
13 Eavesdropping Eavesdropping on Information Set secret key to functions of the IoT to fail and then affect the entire IoT system. An
attack data transmitted leakage filter noise data effective way to solve this issue is to improve the sustainable energy
over a wireless supply performance of edge devices, such as installing solar charging
link
14 Node capture Replace or tamper Important Monitor and
panels. Phishing attacks refer to an attacker who sends phishing web
attack with nodes or information detect malicious sites via email or real-time chat windows (Spaulding & Mohaisen, 2018).
devices in the IoT leakage nodes In this scenario, customers are induced to click on links, and then the
15 Code injection Injecting Attackers Verify the attacker uses them to obtain customers’ personal accounts and pass
attack malicious code control identity of the IoT
words. One of the effective ways to solve this issue is to keep customers
into a node or specific code
device in the IoT functions alert to unfamiliar links. A false data injection attack refers to an
16 Routing Controlling the Data Deploy a secure attacker entering false data in a node or edge device of the IoT (Liu et al.,
information spread of information is routing protocol 2019). These fake data induce the IoT control system to give wrong
attack information by lost or some instructions, which is one of the attacker’s goals for destroying the IoT
manipulating functions fail
routing protocols
system. To solve a false data injection attack, it is necessary to add a data
17 authenticity verification mechanism to IoT nodes.
c) IoT architecture. Service-oriented architecture (SOA) is a key
14
Table 9
Architecture classification of the IoT.
Literature Smart deviceslayer Perceptionlayer Cloudlayer ApplicationLayer Transportlayer
ITU (2003)
Jin et al. (2014)

Shamim & Hossain (2016)
Fang et al. (2014)
Sun and Ansari (2016)
Xu et al. (2014)
Li et al. (2015)
Stojkoska & Trivodaliev (2017)
Lin et al. (2017)
Gubbi et al. (2013)
Atzori et al. (2012)

Rahmani et al. (2017)
Gazis (2017)
Abawajy and Hassan (2017)
Jing et al. (2014)
Islam et al. (2015)
Tsai et al. (2014)
Jia, Feng, Fan, and Lei (2012)
Domingo (2012)
Atzori et al. (2010)
technology for integrating heterogeneous systems or devices and has 6. Conclusion

been successfully applied in the fields of cloud computing and vehicle
networking (Da-Silva, Da-Costa, Crovato, & Righi, 2020; Paulraj, Swa The IoT is permanently changing our world and has been applied in
mynathan, & Madhaiyan, 2012). The key for SOA to be widely used in many fields, including smart healthcare, agriculture and manufacturing.
the IoT is that it supports the construction of a flexible multilayer SOA Since the IoT was introduced, many academics and practitioners have
framework based on specific business needs. For example, the ITU begun to study this disruptive technique. However, there is still a lack of
proposes that an IoT framework should consist of five parts: sensing, a systematic evaluation on this topic from a bibliometric perspective. In
access, middle, network and application layers. Domingo (2012); Li particular, the current literature does not answer the four questions well,
et al. (2013) proposed a more simplified IoT layer framework that in including what is the basic bibliometric overview of the IoT, what are
cludes three parts: perception, network and service layers. Hossain and the collaboration networks of IoT research, what are the thematic trends
Muhammad (2016) proposed an IoT framework that consists of four of IoT development, and what are the main challenges and solutions for
parts: things and devices, communication gateways, clouds and data the IoT? To address this gap, the systematic literature review method
centres, and applications and services. With the aim of exploring the and data sources are described. Finding and insights are then illustrated
composition framework of the IoT, this paper conducts a deep review of through the bibliometric overview, network analysis, thematic trend
20 highly cited papers related to the basic composition framework of the and challenge analysis. Some valuable conclusions of this paper are as
IoT. The composition framework of the IoT are shown in Table 8. follows:
The above statistical results (Table 8) show that the research results
of 20 scholars have given 71 components of the IoT, but there are some 1) Developing economic regions are playing a key role in promoting
overlapping and duplicate parts among them. Therefore, these 71 ele collaborative research on the IoT through their ongoing search for
ments of the IoT are divided into five independent sets in this paper by cooperation with other countries. In addition, institutions in devel
using classification and summary methods, and the independent sets oped economies rarely take the initiative to cooperate with in
include smart device, perception, clouds, transportation and application stitutions in other economies, and they tend to choose institutions
layers. The research status statistics of these five parts of the IoT are that are better than themselves as partners.
shown in Table 9. 2) The mainstream study of the IoT mainly focuses on IoT security,
According to the five independent sets (shown in Table 9), the cur wireless sensor networks, IoT management, IoT challenges, and
rent mainstream IoT system mainly includes five parts: a device layer, privacy. In addition, the keyword thematic evolution shows that
perception layer, cloud layer, transport layer and application layer. This security and algorithm issues have become basic themes in the IoT
IoT architecture has been applied to various fields, including multi- field in recent years.
intelligent control systems (Hadipour, Derakhshandeh, & Shiran, 3) The IEEE Internet of Things Journal is the youngest journal and was
2020), cyber-physical systems (Silva & Jardim-Goncalves, 2019), busi launched in 2014. This journal mainly publishes IoT-related articles.
ness model renewal (Rocha, Fernandes Narcizo, & Gianotti, 2019), and In addition, the IEEE Internet of Things Journal is the most influ
smart manufacturing systems (Jeong, Na, Kim, & Cho, 2018). ential journal, with 3043 citations, followed by IEEE Communica
tions Magazine, with 2468 citations.
15
4) China publishes the most articles and is the main productive country Al-Turjman, F., Nawaz, M. H., & Ulusar, U. D. (2020). Intelligence in the Internet of
Medical Things era: A systematic review of current and future trends. Computer
with 1528 publications, followed by the USA with 896 publications.
Communications, 150, 644–660.
In terms of citation indicators, the USA has the largest global influ Aria, M. (2020). Bilionshiny biliomertrix. https://www.bibliometrix.org/biblioshiny/, 1,
ence, with 21,910 citations. The USA and China have the strongest 1–2.
collaborative relationship, with 270 collaborations, followed by Atzori, L., Iera, A., & Morabito, G. (2010). The Internet of Things: A survey. Computer
Networks, 54(15), 2787–2805.
China and the UK, with 104 collaborations. Beijing University of Atzori, L., Iera, A., & Morabito, G. (2011). SIoT: Giving a social structure to the Internet
Posts and Telecommunications is the most productive institution, of Things. IEEE Communications Letters, 15(11), 1193–1195.
with 101 publications. The University of Cagliari is the most influ Atzori, L., Iera, A., & Morabito, G. (2014). From “smart objects” to “social objects”: The
next evolutionary step of the internet of things. IEEE Communications Magazine, 52
ential institution, with 6526 citations. Guizani, Mohsen is the most (1), 97–105.
influential author in terms of the normalized citation index. Guizani, Atzori, L., Iera, A., Morabito, G., & Nitti, M. (2012). The Social Internet of Things (SIoT) –
Mohsen mainly researches IoT technologies, protocols and applica When social networks meet the Internet of Things: Concept, architecture and
network characterization. Computer Networks, 56(16), 3594–3608.
tions. Atzori, Luigi is the most influential author, with 6501 citations. Aziz, A. A., Ginting, L., Setiawan, D., Park, J. H., Tran, N. M., Yeon, G. Y., Kim, D. I., &
Choi, K. W. (2019). Battery-less location tracking for Internet of Things:
Moreover, this paper designs a general IoT architecture and analyses Simultaneous wireless power transfer and positioning. IEEE Internet of Things Journal,
6(5), 9147–9164.
the challenges that the IoT will face in the future. In particular, trust and Babar, M., & Arif, F. (2019). Real-time data processing scheme using big data analytics in
security challenges and potential solutions to these challenges are ana internet of things based smart transportation environment. Journal of Ambient
lysed and discussed in detail. In terms of TM, determining how to design Intelligence and Humanized Computing, 10(10), 4167–4177.
Ben, S., Olivereau, Y., Zeghlache, D., & Laurent, M. (2013). Trust management system
a more trustworthy IoT system based on the effective fusion of multi-
designs for the Internet of Things: A context-aware and multi-service approach.
source data, qualified service quality and strong user privacy protec Computers & Security, 39, 351–365.
tion has become one of the challenges in the IoT trust field. In terms of Bouzembrak, Y., Klüche, M., Gavai, A., & Marvin, H. J. P. (2019). Internet of Things in
security, sleep deprivation attacks, phishing attacks and false data in food safety: Literature review and a bibliometric analysis. Trends in Food Science &
Technology, 94, 54–64.
jection attacks are discussed in detail. In addition, this paper gives 18 Brussels (2009). Internet of Things— An action plan for Europe. https://www.eesc.euro
common attacks on IoT systems and corresponding solutions to avoid pa.eu/en/ourwork/opinions-information-reports/opinions/internet-things-action-pl
them. The research results of this paper will help relevant researchers, an-europe, 278, 1–15.
Cai, Y., Lu, W., Wang, L., & Xing, W. (2015). Cloud computing research analysis using
entrepreneurs and governments have a clearer understanding of IoT bibliometric method. International Journal of Software Engineering and Knowledge
systems. Engineering, 25(03), 551–571.
Finally, the core difficulty of the IoT is how to realize dynamic Carreño, R., Aguilar, V., Pacheco, D., Acevedo, M. A., Yu, W., & Acevedo, M. E. (2019).
An IoT expert system shell in block-chain technology with ELM as inference engine.
optimization based on real-time data. Therefore, this paper suggests that International Journal of Information Technology & Decision Making, 18(01), 87–104.
researchers of the IoT should pay more attention to the dynamic opti Da-Silva, F. S. T., Da-Costa, C. A., Crovato, C. D. P., & Righi, R. D. (2020). Looking at
mization method of the IoT in various application scenarios in future energy through the lens of Industry 4.0: A systematic literature review of concerns
and challenges. Computers & Industrial Engineering, 143, 21.
research work, which is also one of the key components of this paper for Dang, L. M., Piran, M. J., Han, D., Min, K., & Moon, H. (2019). A survey on internet of
the future. Practitioners should pay more attention to the security and things and cloud computing for healthcare. Electronics, 8(7), 49–59.
reliability of the IoT in the future. Ding, L., Hu, B., Ke, C., Wang, T., & Chang, S. (2019). Effects of IoT technology on gray
market: An analysis based on traceability system design. Computers & Industrial
Engineering, 136, 80–94.
Acknowledgments Domingo, M. C. (2012). An overview of the Internet of Things for people with
disabilities. Journal of Network and Computer Applications, 35(2), 584–596.
This research is funded by the Chongqing Science and Technology Egger, J., & Masood, T. (2020). Augmented reality in support of intelligent
manufacturing – A systematic literature review. Computers & Industrial Engineering,
Commission (Project no. cstc2019jscx-msxmX0189), the National Nat 140, 106195. https://doi.org/10.1016/j.cie.2019.106195
ural Science Foundation of China (72071006 and 62073007), and the Fang, S., Xu, L. D., Zhu, Y., Ahati, J., Pei, H., Yan, J., & Liu, Z. (2014). An integrated
Fundamental Research Funds for the Central Universities, Project No. system for regional environmental monitoring and management based on Internet of
Things. IEEE Transactions on Industrial Informatics, 10(2), 1596–1605.
2018CDXYJX0019. Gazis, V. (2017). A survey of standards for machine-to-machine and the Internet of
Things. IEEE Communications Surveys and Tutorials, 19(1), 482–511.
Appendix A. Supplementary material Ghadimi, P., Wang, C., & Lim, M. K. (2019). Sustainable supply chain modeling and
analysis: Past debate, present problems and future challenges. Resources,
Conservation and Recycling, 140, 72–84.
Appendix A1 shows the institutional co-authorship network. Ap Gibson, E., Daim, T., Garces, E., & Dabic, M. (2018). Technology foresight: A bibliometric
pendix A2 briefly introduces the country co-authorship network. Ap analysis to identify leading and emerging methods. Foresight and STI Governance, 12
(1), 6–24.
pendix A3 presents the authors’ citation network. Appendix A4 shows Gubbi, J., Buyya, R., Marusic, S., & Palaniswami, M. (2013). Internet of Things (IoT): A
the coupling analysis results based on sources. Appendix A5 presents the vision, architectural elements, and future directions. Future Generation Computer
publication coupling results between different countries. Appendix A6 Systems, 29(7), 1645–1660.
Hadipour, M., Derakhshandeh, J. F., & Shiran, M. A. (2020). An experimental setup of
shows the sensitivity analysis of different parameters to clustering re
multi-intelligent control system (MICS) of water management using the Internet of
sults. Appendix A7 presents the bibliographic coupling analysis results. Things (IoT). ISA Transactions, 96, 309–326.
Supplementary data to this article can be found online at https://doi. Harbi, Y., Aliouat, Z., Harous, S., Bentaleb, A., & Refoufi, A. (2019). A review of security
org/10.1016/j.cie.2021.107174. in Internet of Things. Wireless Personal Communications, 108(1), 325–344.
Hossain, M. S., & Muhammad, G. (2016). Cloud-assisted Industrial Internet of Things
(IIoT) – Enabled framework for health monitoring. Computer Networks, 101,
References 192–202.
Hu, G., Wang, L., Ni, R., & Liu, W. (2020). Which h-index? An exploration within the web
Abawajy, J. H., & Hassan, M. M. (2017). Federated internet of things and cloud of science. Scientometrics, 123(3), 1225–1233.
computing pervasive patient health monitoring system. IEEE Communications Husain, A. J., & Mohamed, M. A. M. (2019). IMBF - counteracting denial-of-sleep attacks
Magazine, 55(1), 48–53. in 6LowPAN based Internet of Things. Journal of Information Science and Engineering,
Addo-Tenkorang, R., & Helo, P. T. (2016). Big data applications in operations/supply- 35(2), 361–374.
chain management: A literature review. Computers & Industrial Engineering, 101, Islam, S. M. R., Kwak, D., Kabir, M. H., Hossain, M., & Kwak, K. S. (2015). The internet of
528–543. things for health care: A comprehensive survey. IEEE Access, 3, 678–708.
Airehrour, D., Gutierrez, J. A., & Ray, S. K. (2019). SecTrust-RPL: A secure trust-aware ITU. (2003). Internet of things reports. http://www.itu.int/osg/spu/publications/intern
RPL routing protocol for Internet of Things. Future Generation Computer Systems, 93, etofthings/, 1, 1–5.
860–876. Jeong, S., Na, W., Kim, J., & Cho, S. (2018). Internet of things for smart manufacturing
Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M., & Ayyash, M. (2015). Internet system: trust issues in resource allocation. IEEE Internet of Things Journal, 5(6),
of Things: A survey on enabling technologies, protocols, and applications. IEEE 4418–4427.
Communications Surveys and Tutorials, 17(4), 2347–2376.
16
Jia, X., Feng, Q., Fan, T., & Lei, Q. (2012). RFID technology and its applications in Roman, R., Zhou, J., & Lopez, J. (2013). On the features and challenges of security and
Internet of Things (IoT). In International conference on consumer electronics, privacy in distributed internet of things. Computer Networks, 57(10), 2266–2279.
communications and networks (pp. 1282–1285). Santoro, G., Vrontis, D., Thrassou, A., & Dezi, L. (2018). The Internet of Things: Building
Jin, J., Gubbi, J., Marusic, S., & Palaniswami, M. (2014). An information framework for a knowledge management system for open innovation and knowledge management
creating a smart city through internet of things. IEEE Internet of Things Journal, 1(2), capacity. Technological Forecasting and Social Change, 136, 347–354.
112–121. Schoenberger, C. R. (2002). The Internet of things. Forbes, 169(6), 155–185.
Jing, Q., Vasilakos, A. V., Wan, J., Lu, J., & Qiu, D. (2014). Security of the Internet of Shin, D. (2014). A socio-technical framework for Internet-of-Things design: A human-
Things: Perspectives and challenges. Wireless Networks, 20(8), 2481–2501. centered design for the Internet of Things. Telematics and Informatics, 31(4),
Kwak, J., Cho, C., Shin, Y., & Yang, S. (2020). IntelliTC: Intelligent inter-DC traffic 519–531.
controller for the Internet of everything service based on fog computing. IET Silva, E. M., & Jardim-Goncalves, R. (2019). Cyber-Physical Systems: A multi-criteria
Communications, 14(2), 193–205. assessment for Internet-of-Things (IoT) systems. Enterprise Information Systems, 2019,
Lee, D., & Lee, H. (2018). IoT service classification and clustering for integration of IoT 1698060.
service platforms. Journal of Supercomputing, 74(12), 6859–6875. Sinha, A., Shrivastava, G., & Kumar, P. (2019). Architecting user-centric internet of
Lee, I., & Lee, K. (2015). The Internet of Things (IoT): Applications, investments, and things for smart agriculture. Sustainable Computing: Informatics and Systems, 23,
challenges for enterprises. Business Horizons, 58(4), 431–440. 88–102.
Li, D., Cai, Z., Deng, L., Yao, X., & Wang, H. H. (2019). Information security model of Small, H. (1973). Cocitation in scientific literature - new measure of relationship
block chain based on intrusion sensing in the IoT environment. Cluster Computing, 22 between 2 documents. Journal of the American Society for Information Science, 24(4),
(S1), 451–468. 265–269.
Li, H., & Zhou, X. (2011). Study on security architecture for Internet of Things. Applied Spaulding, J., Mohaisen, A., & IEEE. (2018). Defending internet of things against
Informatics and Communication, 224, 404–425. malicious domain names using D-FENS. New York: IEEE. 10, 387–392.
Li, M. (2017). Internet of people: editorial. Concurrency and Computation-Practice & Stojkoska, B. L. R., & Trivodaliev, K. V. (2017). A review of Internet of Things for smart
Experience, 29(3), e4050. https://doi.org/10.1002/cpe.v29.310.1002/cpe.4050 home: Challenges and solutions. Journal of Cleaner Production, 140, 1454–1464.
Li, Q., Wang, Z.-Y., Li, W.-H., Li, J., Wang, C., & Du, R.-Y. (2013). Applications Su, J. F., Bai, Q., Sindakis, S., Zhang, X. F., & Yang, T. (2020). Vulnerability of
integration in a hybrid cloud computing environment: Modelling and platform. multinational corporation knowledge network facing resource loss A super-network
Enterprise Information Systems, 7(3), 237–271. perspective. Management Decision, 20, 2019-0227.
Li, S., Xu, L. D., & Zhao, S. (2015). The internet of things: A survey. Information Systems Su, J. F., Yang, Y., & Yang, T. (2018). Measuring knowledge diffusion efficiency in R&D
Frontiers, 17(2), 243–259. networks. Knowledge Management Research & Practice, 16(2), 208–219.
Li, X., Lu, R., Liang, X., Shen, X., Chen, J., & Lin, X. (2011). Smart community: An Sun, W., Cai, Z., Li, Y., Liu, F., Fang, S., & Wang, G. (2018). Security and Privacy in the
internet of things application. IEEE Communications Magazine, 49(11), 68–75. Medical Internet of Things: A Review. Security and Communication Networks, 2018,
Li, Y., Lim, M. K., Tan, Y., Lee, S. Y., & Tseng, M.-L. (2020). Sharing economy to improve 1–9.
routing for urban logistics distribution using electric vehicles. Resources, Conservation Sun, X., & Ansari, N. (2016). EdgeIoT: Mobile edge computing for the internet of things:
and Recycling, 153, 104585. https://doi.org/10.1016/j.resconrec.2019.104585 IEEE Press, 12, 22–29.
Lim, M. K., & Jones, C. (2017). Resource efficiency and sustainability in logistics and Tsai, C.-W., Lai, C.-F., Chiang, M.-C., & Yang, L. T. (2014). Data mining for internet of
supply chain management. International Journal of Logistics Research and Applications, things: A survey. IEEE Communications Surveys and Tutorials, 16(1), 77–97.
20(1), 20–21. Tseng, M.-L., Lim, M. K., & Wu, K.-J. (2019). Improving the benefits and costs on
Lim, M. K., Tseng, M.-L., Tan, K. H., & Bui, T. D. (2017). Knowledge management in sustainable supply chain finance under uncertainty. International Journal of
sustainable supply chain management: Improving performance through an Production Economics, 218, 308–321.
interpretive structural modelling approach. Journal of Cleaner Production, 162, Tu, M. R., Lim, M. K., & Yang, M. F. (2018a). IoT-based production logistics and supply
806–816. chain system - Part 1 Modeling IoT-based manufacturing IoT supply chain. Industrial
Lim, M. K., Wang, J. X., Wang, C., & Tseng, M. L. (2020). A novel method for green Management & Data Systems, 118(1), 65–95.
delivery mode considering shared vehicles in the IoT environment. Industrial Tu, M., Lim, M. K., & Yang, M.-F. (2018b). IoT-based production logistics and supply
Management & Data Systems, 120(9), 1733–1757. chain system – Part 2: IoT-based cyber-physical system: A framework and
Lim, M. K., Xiong, W. Q., & Lei, Z. M. (2020). Theory, supporting technology and evaluation. IMDS, 118(1), 96–125.
application analysis of cloud manufacturing: a systematic and comprehensive Wang, C., Ghadimi, P., Lim, M. K., & Tseng, M.-L. (2019). A literature review of
literature review. Industrial Management & Data Systems, 120(8), 1585–1614. sustainable consumption and production: A comparative analysis in developed and
Lin, J., Yu, W., Zhang, N., Yang, X., Zhang, H., & Zhao, W. (2017). A survey on internet of developing economies. Journal of Cleaner Production, 206, 741–754.
things: Architecture, enabling technologies, security and privacy, and applications. Wang, C., Lim, M. K., & Lyons, A. (2019). Twenty years of the International Journal of
IEEE Internet of Things Journal, 4(5), 1125–1142. Logistics Research and Applications: a bibliometric overview. International Journal of
Liu, K., Bi, Y., & Liu, D. (2020). Internet of Things based acquisition system of industrial Logistics-Research and Applications, 22(3), 304–323.
intelligent bar code for smart city applications. Computer Communications, 150, Wang, C., Lim, M. K., Zhao, L. F., Tseng, M. L., Chien, C. F., & Lev, B. (2020). The
325–333. evolution of Omega-The International Journal of Management Science over the past
Liu, X., Qian, C., Hatcher, W. G., Xu, H., Liao, W., & Yu, W. (2019). Secure Internet of 40 years: A bibliometric overview. Omega-International Journal of Management
Things (IoT)-Based smart-world critical infrastructures: Survey, case study and Science, 93, 102098.
research opportunities. IEEE Access, 7, 79523–79544. Wang, C., Zhao, L. F., Vilela, A. L. M., & Lim, M. K. (2019). The evolution of Industrial
Manavalan, E., & Jayakrishna, K. (2019). A review of Internet of Things (IoT) embedded Management & Data Systems over the past 25 years A bibliometric overview.
sustainable supply chain for industry 4.0 requirements. Computers & Industrial Industrial Management & Data Systems, 119(1), 2–34.
Engineering, 127, 925–953. Wang, J. X., Lim, M. K., Zhan, Y. Z., & Wang, X. F. (2020). An intelligent logistics service
Most. (2012). The 12th five-year development plan for the Internet of things. http system for enhancing dispatching operations in an IoT environment. Transportation
://www.gov.cn/zwgk/2012-02/14/content_2065999.htm, 1(1), 1–5. Research Part E-Logistics and Transportation Review, 135, 23.
Muñuzuri, J., Onieva, L., Cortés, P., & Guadix, J. (2020). Using IoT data and applications Xu, L. D., He, W., & Li, S. (2014). Internet of things in industries: A survey. IEEE
to improve port-based intermodal supply chains. Computers & Industrial Engineering, Transactions on Industrial Informatics, 10(4), 2233–2243.
139, 105668. https://doi.org/10.1016/j.cie.2019.01.042 Yan, L., & Wang, K. (2010). Trust control in heterogeneous networks for Internet of
Nees J. V. E., & Waltman, L. (2020). VOSviewer visualizing scientific landscapes. https:// Things. In International Conference on Computer Application & System Modeling, 1 pp.
www.vosviewer.com/, 1, 1–8. 12–22). IEEE.
Ning, H., Liu, H., & Yang, L. T. (2013). Cyberentity security in the internet of things. Yan, Z., Zhang, P., & Vasilakos, A. V. (2014). A survey on trust management for Internet
Computer, 46(4), 46–53. of Things. Journal of Network and Computer Applications, 42, 120–134.
Nitti, M., Girau, R., & Atzori, L. (2014). Trustworthiness management in the social Zanella, A., Bui, N., Castellani, A., Vangelista, L., & Zorzi, M. (2014). Internet of things
internet of things. IEEE Transactions on Knowledge and Data Engineering, 26(5), for smart cities. IEEE Internet of Things Journal, 1(1), 22–32.
1253–1266. Zhang, N., Yang, Y., Su, J. F., & Zheng, Y. J. (2019). Modelling and analysis of complex
Panetta, K. (2016). Market Research Report for internet of things. https://www.gartner. products design based on supernetwork. Kybernetes, 48(5), 861–887.
com/smarterwithgartner/7-technologies-underpin-the-hype-cycle-for-the-internet- Zhang, N., Yang, Y., Wang, J. X., Li, B. D., & Su, J. F. (2018). Identifying core parts in
of-things-2016/, 3(1), 12–19. complex mechanical product for change management and sustainable design.
Paulraj, D., Swamynathan, S., & Madhaiyan, M. (2012). Process model-based atomic Sustainability, 10(12), 15.
service discovery and composition of composite semantic web services using web Zhang, N., Yang, Y., Zheng, Y. J., & Su, J. F. (2019). Module partition of complex
ontology language for services (OWL-S). Enterprise Information Systems, 6(4), mechanical products based on weighted complex networks. Journal of Intelligent
445–471. Manufacturing, 30(4), 1973–1998.
Rahmani, A. M., Gia, T. N., Negash, B., Anzanpour, A., Azimi, I., Jiang, M., & Liljeberg, P. Zhang, Y., & Wen, J. T. (2017). The IoT electric business model: Using blockchain
(2018). Exploiting smart e-Health gateways at the edge of healthcare Internet-of- technology for the internet of things. Peer-to-Peer Networking and Applications, 10(4),
Things: A fog computing approach. Future Generation Computer Systems, 78, 641–658. 983–994.
Rocha, C., Fernandes Narcizo, C., & Gianotti, E. (2019). Internet of management artifacts: Zhang, Y., & Zhao, W. J. (2019). Dalian Institute of Chemical Physics, Chinese Academy
Internet of Things Architecture for business model renewal. International Journal of of Sciences. National Science Review, 6(4), 843.
Innovation and Technology Management, 16(08), 1950062. https://doi.org/10.1142/
S0219877019500627
17

1 s2.0 S0360835221000784 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0360835221000784 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0360835221000784 Main

Uploaded by

Copyright:

Available Formats

Computers & Industrial Engineering 155 (2021) 107174

Contents lists available at ScienceDirect

Computers & Industrial Engineering

a) Research articles b) Review articles

Table 3 Telecommunications, with 97 publications. It is worth mentioning that

Fig. 5. The source citation network. Note:

1 Atzori, Luigi 80 403 14 6529 34.13 2014 466.36 2.44

Table 5 4.3. Bibliographic coupling network analysis

Table 6 5.1. Keyword co-occurrence network analysis

To reveal the evolution of keyword trends, strategic diagrams are

Table 7 Table 7 (continued )

Jin et al. (2014)

Fang et al. (2014)

Sun and Ansari (2016)

Lin et al. (2017)

Gubbi et al. (2013)

Atzori et al. (2012)

Abawajy and Hassan (2017)

Jing et al. (2014)

Islam et al. (2015)

Tsai et al. (2014)

Jia, Feng, Fan, and Lei (2012)

Atzori et al. (2010)

technology for integrating heterogeneous systems or devices and has 6. Conclusion

You might also like