https://doi.or g/10.31449/inf.v47i7.4783 Informatica 47 (2023) 91–1 14 91 An Exploratory Bibliometric Analysis of the L iteratur e on Age of Information-A war e Unmanned Aerial V ehicles Aided Communication Umar Ali Bukar 1 , Md Shohel Sayeed 1∗ , Siti Fatimah Abdul Razak 1 , Sumendra Y ogarayan 1 and Oluwatosin Ahmed Amodu 2, 3 1 Centre for Intelligent Cloud Computing (CICC), Faculty of Information Science T echnology , Multimedia University , Melaka, Malaysia. 2 Department of Electrical, Electronics & Systems Engineering, Universiti Kebangsaan Malaysia (UKM), 43600, Bangi, Selangor , Malaysia. 3 Information and Communication Engineering Department, Elizade University , Ilara-Mokin, Ondo State, Nigeria. Email: umarfalmata@gmail.com, shohel.sayeed@mmu.edu.my ∗ , fatimah.razak@mmu.edu.my , sumendra@mmu.edu.my , amodu_o_a@ieee.or g Overview paper Keywords: UA V , age of information, bibliometric analysis, information freshness, visualization, text analysis Received: Apr 5, 2023 Real-time status updates r equir e mor e fr equent updates with fr esh information. This study investigates the applications and r esear ch potential of unmanned aerial vehicles (UA V) for achieving information fr eshness in time-critical applications to emphasize important aspects of this subject based on a thor ough statistical analysis of curr ent r esear ch tr ends. Particularly using the Scopus database, a bibliometric analysis is con- ducted on 122 articles written in English and published between 2018 and 2023. This analysis pr ovides a knowledge map of past r esear ch on this subject and the journey so far , especially concerning major subjects, patterns of citations, publication activities, and the state of cooperation among contributors thr oughout the UA V -information fr eshness r esear ch history . Accor ding to the findings, applying various methods, such as deep r einfor cement learning and optimization algorithms, has been evident. In contrast, ener gy efficiency and harvesting, trajectory planning and design, and scheduling ar e issues attracting r esear chers’ inter est. Finally , the study offers implications and r ecommendations such as fostering inter disciplinary collabo- ration, furthering and impr oving on DRL and optimization algorithms, addr essing ener gy efficiency and harvesting, enhancing trajectory planning and design, emphasizing scheduling strategies, and bridging the gap between r esear ch and practice. Povzetek: Opravljena je bila bibliometrična analiza 122 člankov o br ezpilotnih letalnikih, objavljenih med letoma 2018 in 2023, z namenom pridobivanja osveženih informacij. 1 Intr oduction The next-generation wireless networks need to handle a wider variety of services than their predecessors. They should provide communications with low latency , high reli- ability , and improved mobile broadband. Unmanned aerial vehicles, often known as UA Vs, can of fer workable solu- tions for next-generation networks, and they can also serve as aerial base stations (BS) for data collection and trans- mission [ 1 ]. UA Vs have basically evolved as a form of cutting-edge technology that is relatively inexpensive, mo- bile, flexible, and able to communicate with terrestrial in- frastructures via direct line of sight (LoS) links [ 2 ]. It has the advantages of flexible deployment and maneuverability , as well as low cost. The UA V is one of the most promising technologies and has become an interesting topic in indus- try and academia to drive the development of Internet of Things (IoT) applications [ 3 , 4 , 5 , 6 ]. On the other hand, a concept known as the ”Age of Infor - mation” (AoI) has been developed to determine the fresh- ness of data, especially in real-time and (time-critical) ap- plications [ 7 ]. Likewise, AoI has been designed and re- searched in numerous fields as a metric for assigning tem- poral values to information aging [ 8 ]. For example, the AoI is a critical metric in the data aggregation and analytics for IoT [ 9 ]. AoI is the time elapsed since generating the lat- est received packet at the data point; the freshness metric in status update systems [ 9 ]. Accordingly , AoI is the amount of time that has passed from the creation of information. Thus, the information generated long ago is characterized by a greater AoI, indicating that its current condition may diver ge from what is expected [ 10 ]. Simply put, AoI is the measurement of the amount of time that has passed since the generation of the most recent update that was received about a process. It is an essential metric in networks such as IoT , mainly when the application requires up-to-date in- 92 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. Figure 1: Illustration depicting some UA Vs applications formation [ 7 ]. Nonetheless, the evolution of UA Vs has shown promise in a variety of applications of wireless communication due to the UA Vs’ high coverage, promising rates, and flexible installation [ 1 1 ]. The UA V possesses vast abilities such as high flight ef ficiency on fixed routes and data collection on the move [ 12 ]. When the UA V returns, the data collected by the ground fixed-point equipment while in the fixed-point cruising mode can be obtained from the multi-source sensor equipment mounted on the UA V in complex mountain en- vironments. The data can then be stored in the data center for additional processing and analysis [ 12 ]. Additionally , when it comes to maintaining some national forest parks and keeping an eye on the area close to cities, UA V pa- trol has gradually replaced manual patrol as the preferred method. A typical UA V application is demonstrated in Fig. 1 [ 13 ], which shows UA V deployment in three dif ferent ap- plications. Specifically , Fig. 1 a depicts disaster or emer - gency applications, Fig. 1 b shows UA V applications in agricultural farms to collect data from crops, and Fig. 1 c represents the applications in industrial IoT . In the applications (Fig. 1 b and 1 c), the sensors relay in- formation to the BS (i.e., data center) via the UA V for trans- mission. The UA V flies and collects data from at least one sensor , communicates with other data points, then returns to a data center to transmit the data [ 14 ]. A UA V is deployed to collect and transmit the sensor data to the data center , es- pecially in agriculture and industrial applications. The data collection occurs via wireless transmission medium when- ever the UA V is in close vicinity to the sensor . The UA V can be equipped with a variety of sensors to gather data, including changes in speed, temperature, light, distance, chemical signals, wind, and sound. It also can detect the existence of magnetic objects. Hence, UA Vs provide a plat- form for data collection that quickly collects information from broad perspectives. The capability of UA V camera can be used to create precise three-dimensional representa- tions of locations or objects, especially for disaster moni- toring and surveillance (see Fig. 1 a). Moreover , the UA V channel exhibits several exclusive properties like shadow- ing, 3D deployment, high mobility , and spatial and tempo- ral non-stationarity [ 15 , 16 ]. According to [ 17 ], the UA V channel may be divided into two distinct categories: air - to-air and air -to-ground. Small to medium-sized manned aircraft and UA Vs are the two forms of aerial vehicles uti- lized for channel measurements. Channel measurements of the former are costly , whereas channel measurements of the latter have the potential to save costs significantly [ 15 ]. Using the VOSviewer software, this study conducted a bibliometric analysis of 122 papers published in En- glish between 2018 and 2023 and sourced from the Scopus database. The study aims to analyze bibliometric data re- garding UA V and AoI. The findings of fer a schematic rep- resentation of the information generated and disseminated by earlier research. It provides insights into significant sub- jects, citation trends, publication activities, collaboration status among contributors, and aggregated UA V research contributions in disaster management scenarios. The remainder of this paper is structured as follows: Sec- tion 2 discusses the related reviews and rationale for con- ducting this work; Section 3 provides an explanation of the methodology that was applied while conducting the study; Section 4 of fers a description of the results, which is cov- ered based on keyword analysis, document sources, and ci- tations patterns; Section 5 presents the discussion of the study , which also covers practical perspectives as well as limitations and consideration for future research. Finally , the conclusions drawn from this study is covered in Sec- tion 6 . An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 93 2 Related works Review studies combining the diverse aspects of UA V for AoI-sensitive applications are limited in the current litera- ture [ 18 ]. However , several review ef forts have been made to address potential research areas in UA V communica- tions. For instance, the work by [ 19 ] conducted reviews on UA V path planning and 5G communications. In [ 20 ], the study performed a taxonomic review on UA Vs with routing and trajectory optimization problems. In addition, the work by [ 21 ] reviewed IoT and UA V applications and commu- nication technologies for sustainable smart farming, while [ 22 ] conducted a comprehensive review on ener gy sources for UA Vs. Nevertheless, [ 23 ] succinctly captured various review papers related to using UA Vs for data gathering in IoT applications to minimize AoI. Although the study cov- ers a wide range of aspects related to AoI minimization, there were no bibliometric analysis-based insights giving first-hand detail on the analysis of keywords, document sources and types, most active source titles, geometric dis- tribution of publications, most influential institutions, cita- tions, and textual titles, which provide insights into dif fer - ent trendsetters and some of the statistical analysis-based trends in this field. However , there have been various bibliometric analyses conducted on UA Vs and their related applications. For in- stance, studies have examined UA V usage in dif ferent do- mains, such as UA V and forest [ 24 ], UA V in agriculture and forestry [ 25 ], UA V and precision agriculture and viti- culture [ 26 ], UA V and social network [ 27 ], UA V swarms [ 28 ], UA V and crop monitoring in smallholder farms [ 29 ], drone research and scholarly output [ 30 ], UA V usage in ar - chitecture and urbanism [ 31 ], drone and blast-induced fly- rock [ 32 ], UA V and wheat crop [ 33 ], UA V and cellular net- work [ 34 ], UA V and digitalization of public administration [ 35 ], and drone delivery systems [ 36 ]. Hence, the lack of bibliometric studies focusing on UA V and AoI justifies the need for additional research in this field. In our ef fort to identify existing review studies concerning UA V and AoI, T able 1 presents the few papers identified from the litera- ture. The summary of these studies (refer to T able 1 ) pro- vides important insights that justify the use of bibliometric analysis in this area. Furthermore, T able 2 presents a critical analysis of the existing review papers combining UA V and AoI to demon- strate the rationale explaining what is missing in the cur - rent literature and how this study remedies these gaps. The study of related literature reveals that there is a lack of bib- liometric studies on this subject which could help new re- searchers identify research clusters, as well as trends across dif ferent sub-domains. However , bibliometric analysis is a valuable approach for examining the research landscape and understanding the current state of knowledge in a spe- cific field [ 38 , 39 , 40 ]. In the context of UA Vs and AoI, sev- eral studies have explored dif ferent aspects of this domain [ 37 , 23 , 18 ], highlighting the need for a comprehensive bib- liometric analysis to synthesize and analyze the existing lit- erature. By conducting a bibliometric analysis, this study aims to address the limitations of existing reviews and pro- vide a comprehensive overview of the research trends, in- fluential studies, and emer ging themes in the field of UA Vs and AoI. This analysis will enable researchers and practi- tioners to identify gaps, highlight high-impact areas, and guide future investigations in this dynamic and evolving do- main. 3 Methodology The bibliometric and systematic approaches to literature review were used to draw inferences from the research and meta-data on this subject to ensure the methodology aligns with the laid-out research objectives. In this context, VOSviewer bibliometric program was adopted to analyze the bibliographic data obtained [ 41 ]. The bibliographic analysis involves the use of citations as the variable of interest. Citation analysis is a bibliometric approach based on the premise that citations can be used as indicators of activity within a scientific field [ 42 , 43 ]. This means a frequently cited article is relatively more signifi- cant in the field under study [ 44 ]. The authors in [ 45 ] added that citation data could be used to identify the most influ- ential papers, both ”locally” (within a field) and ”globally” (among the entire research community). Citation analysis facilitates the discovery of important research streams as well as attribution, access, use, management, and retrieval of scholarly content [ 46 ]. Additionally , the quantitative research methodology is adopted in this work. Several databases, such as Google Scholar , W eb of Science (W oS), and Scopus, provide cita- tion information. While Google Scholar is more inclusive, W oS and Scopus are more selective in terms of t he quality of the journals they capture. Building on the research tech- niques and procedures used in previous bibliometric studies (see [ 45 , 47 , 48 ], this study adapted the concept deployed in [ 47 ]. Moreover , Scopus was selected as the primary data source for this analysis for several reasons, including credi- bility and peer review . It also covers most W oS papers and has a wide range of references, abstracts, and summaries under accepted practices [ 49 ]. Additionally , more than 87 million papers and 25,000 ac- tive titles are available on Scopus. The Scopus database has the most comprehensive coverage of abstract and citation databases that spans multiple disciplines. The database is a trustworthy resource for obtaining global academic knowl- edge and is regularly updated [ 47 ]. It is also hard to dis- regard the Scopus h-index tool, which determines the dif- ferent category of book, author , or journal [ 50 ]. The article database search and selection was created using a five-step process [ 47 ], as shown in Fig. 2 . Accordingly , this study analyzes 122 articles obtained between 2018 and 2023. The selection o f the articles was applied exclusively based on the language used, English, and identifying contents related to AoI and UA V . 94 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. T able 1: Summary and focus of existing review papers Ref Y ear Focus Summary Method [ 37 ] 2018 DDDAS with an aspect of UA V and AoI This paper provides a comprehensive review of Dynamic Data Driven Applications Systems (DDDAS), a systems design framework that integrates physical model simulations, real- time measurements, statistical methods, and computation archi- tectures. It highlights the successes of DDDAS in various fields such as natural disaster assessment, space awareness, UA V de- sign, and biomedical applications. The paper also discusses recent developments in DDDAS related to information man- agement architecture, sensor design, information filtering, and computational systems. Descriptive review [ 18 ] 2023 UA V , AoI, WSNs, and IoT This article provides a comprehensive review of 20 selected articles on minimizing the Age of Information (AoI) in UA V - assisted data gathering for wireless sensor networks (WSNs) and Internet of Things (IoT) applications. It explores various techniques, including machine learning and optimization meth- ods, to optimize UA V trajectory , scheduling, and ener gy source acquisition, and discusses the challenges, lessons learned, and future directions in this field. Systematic [ 23 ] 2023 UA V , AoI, WSNs, and IoT This article presents a systematic literature review (SLR) on age minimization in UA V -assisted data-gathering architectures for WSNs and IoT networks. The review identifies three cru- cial design aspects: ener gy management, flight trajectory , and UA V/sensor node scheduling, and discusses various issues and considerations related to these aspects, as well as future direc- tions for optimization and system improvements. Systematic This study UA V , AoI, and other paradigms. Knowledge map concerning major subjects, patterns of ci- tations, publication activities, and the state of cooperation among contributors throughout the UA V -information freshness research literature. Bibliometric T able 2: Critical analysis of existing reviews and research gap Ref Y ear DDDAS UA V AoI WSNs IoT Descriptive SLR Bibliometric [ 37 ] 2018 [ 18 ] 2023 [ 23 ] 2023 This study Several publications, including IEEE Internet of Things Journal, IEEE T ransactions on vehicular technology , IEEE T ransactions on wireless communication, and IEEE T rans- actions on Communications, published technical studies on UA V -assisted AoI reduction. V ersion 1.6.18 of the VOSviewer program is used in this study to analyze the bibliometric data. The tool is an open-source program for designing and developing bibliometric networks [ 41 ]. The bibliometric data was then processed to produce a file with the structure and format needed for network analysis. A visualization map was developed based on the bibliometric data obtained from Scopus to gain a better and more insight- ful understanding of the bibliometric findings related to the study themes. The program is integrated with a text-mining feature, making it attractive to researchers. Given this, a lar ge cor - pus of scholarly literature has made use of the software to build co-occurrence visualization maps related to the subjects studied [ 47 , 48 ]. Hence, the research methodol- ogy was divided mainly into three stages: (1) gathering and assessing the relevant materials, (2) analyzing the bib- liographic data, and (3) Extracting information based on the main keywords identified in previous stages. The ini- tial step involved studying the collected papers using the following procedure: the Scopus database was searched, quantitative analysis was performed, additional searches focusing on disaster applications were conducted, and data- gathering methods were developed. Fig. 2 shows the method used for gathering data; these steps were done to guarantee accuracy . In the first stage, academic literature publications from the Scopus database were examined in order to highlight An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 95 Figure 2: Research Methodology and Design of Paper Selection and Performance Analysis and classify the primary research trends in the area. At this point, there were no limitations on the range of the publica- tion period. The final articles were selected after conduct- ing title and abstract reading, and the range between 2018 and 2023 was finalized, which is the range of 122 articles. T o find pertinent results, a search employing combinations of two strings was carried out, and the result was utilized according to the purpose of each stage under the data col- lection and processing phase, as shown below . – Search string 1: (”data freshness” OR ”information freshness” OR ”age of information” OR AoI) AND (”unmanned aerial vehicle” OR ”UA V” OR ”drone”). – Search string 2: (”data freshness” OR ”informa- tion freshness” OR ”age of information”) AND (”un- manned aerial vehicle” OR ”UA V” OR ”drone”). The two-search string was designed to of fer more rel- evant and accurate articles on the subject area. The first search string result identifies 404 papers. By critically eval- uating the abstracts of the papers, it was observed that AoI is also a short form for “area of interest”. In the second search string, 327 papers were identified from which 321 papers that are written in the English language were selected. Nev- ertheless, title and abstract reading were conducted on the 321 articles. Finally , the study included 122 for the biblio- metric analysis. 4 Results The search and selection of articles produced 122 papers related to UA V and AoI. The discussion of the results is presented in this section. The bibliometric aspect of the ar - ticles is presented. Accordingly , the study provides discus- sions and information regarding a quantitative study that was conducted on the selected papers. Using the keywords ”Age of information” and ”Unmanned aerial vehicle” iden- tified previously , a bibliometric analysis conducted previ- ously , this study carried out a bibliometric study accord- ing to the published data: articles keywords, article sources, and types, year of publications, publication country , authors institution, and contributions, and journal titles. Accord- ingly , Scopus was used to acquire the bibliographic data utilized in this study . T able 3: Popular authors keywords Author keywords Frequency Percent Age of information 77 26.37 Unmanned aerial vehicle 63 21.58 IoT 16 5.48 Deep reinforcement learning (DRL) 19 6.51 Data collection 15 5.14 T rajectory 14 4.79 T rajectory optimization 9 3.08 T rajectory planning 9 3.08 W ireless sensor network 9 3.08 Scheduling 8 2.74 Optimization 7 2.4 Autonomous aerial vehicle 7 2.4 T ask analysis 7 2.4 Ener gy ef ficiency 6 2.05 Sensors 6 2.05 Convex optimization 5 1.71 Path planning 5 1.71 T rajectory design 5 1.71 Multi-agent deep reinforcement learning 5 1.71 4.1 Keyword analysis Utilizing the text-mining algorithm of the VOSviewer 1.6.18 [ 41 ], the study visualized the keyword information produced by various publications. Several bibliometric re- search has validated this approach [ 51 , 52 , 53 , 48 , 47 ]. Thus, the text-mining technique generates a map that thor - oughly interprets the distance between terms as an indica- 96 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. Figure 3: Map of co-occurrences of authors keywords (See T able 3 for more details). tion of the correlation between various keywords. Accord- ingly , the greater the distance between two or more key- words, the greater the significance of the related terms. The co-occurrences of words in publications were analyzed to determine their interconnectedness based on a unit of anal- ysis (author keyword) and counting method (fractional). Specifically , the network analysis of the author keywords includes only those terms that appear in the database at least five (5) times. These are presented in T able 3 . This study examined the keyword occurrence carefully to ensure the accuracy of the data. 26 keywords out of 292 are deemed suitable for analysis. Then, the duplicates or keywords with synonyms (e.g., ’UA Vs’, ‘UA V’, ‘Unmanned aerial vehi- cle’ etc.) were eliminated. Moreover , the keywords that did not adequately describe anything (e.g., algorithm, anal- ysis, etc.) were also eliminated. Fig. 3 depicts the visualization network and map of the authors’ keywords. It highlights the most used terms in the existing studies through a conceptual map to illustrate the connection between keywords used by the authors [ 41 ]. The keyword size is determined entirely by their presence in the selected articles. According to the result, the major key- word was ’Age of Information’, which was used constantly throughout the study period. Similarly , the terms such as UA V , deep reinforcement learning (DRL), IoT , trajec- tory , data collection, optimization, wireless sensor network (WSN), and scheduling were frequently observed in the lit- erature. Furthermore, the depiction of the keywords as well as co-occurrence (co-word estimation) demonstrates a well- known issue from the literature for AoI-aware UA V deploy- ment. It is vital to consider color -matching of the terms ’age of information’, ’DRL (the blue circle located in the cen- ter with label)’, ‘data collection and trajectory’, and ‘con- vex optimization’. This co-occurrence measure quantifies the strength of the interaction between the terms, particu- larly between UA V technology , its associated issues, and deployed methods. Accordingly , T able 3 displays the most prominent terms that have been utilized by numerous re- searchers in the past. T able 4: Y ear of publications Y ear Frequency Cumulative percent(N=122) 2018 2 1.64 2019 12 9.84 2020 19 15.57 2021 35 28.69 2022 53 43.44 2023 1 0.82 T otal 122 100 4.2 Document sour ce and types The document and source types analysis shows that, within the 122 papers that were chosen as the sample, the con- ference papers made up 50.82% of the total, making them slightly more than the journal articles which constitute 47.54%. Moreover , when it comes to the source and ori- gin of the papers obtained for the analysis, journal sources account for a total of 50%, slightly higher than conference An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 97 Figure 4: Document by year (See also T able 4 ). T able 5: Document type and sources Document type Source type Document type Frequency %(N=122) Source type Frequency %(N=122) Article 58 47.54 Journals 61 50 Conference paper 62 50.82 Conference proceedings 58 47.54 Conference review 0 0 Lecture notes 2 1.64 Book chapter 1 0.82 Book 1 0.82 Editorial 1 0.82 T rade publications 0 0 proceedings sources (47.54%). T able 5 presents the num- ber of all the dif ferent types of papers considered for this study , such as lecture notes, book chapters, and editorials. 4.3 Publications years of the published studies The progression of published studies relating to the topic can be seen in Fig. 4 , which covers the years 2018 through 2023. The study observes a slow but steady rise in the number of publications on AoI-aware UA V deployment. There were only two publications in 2018. The publications started to emer ge more in 2019, which saw the fewest num- ber and subsequently led to a sur ge in the number of publi- cations between 2021 and 2022. This suggests that schol- ars are becoming increasingly interested in UA Vs and AoI. T able 4 contains a comprehensive overview of the years in which research studies were published. The result re- veals that 2022 has the highest number of publications. By observing this graph, the interest in UA V and information freshness is growing which could suggest that more publi- cations will be produced in the future. 4.4 Most active sour ce titles T able 6 provides a summary of the leading and most ac- tive journals that have published works connected to AoI and UA V , with at least 3 publications. The major source title is the IEEE Internet of Things journal with eleven documents. Others include IEEE T ransactions on vehic- ular technology , IEEE transactions on wireless communi- cation, IEEE T ransactions on intelligent transportation sys- tems, IEEE vehicular technology conference, international conference on Communications, and IEEE Journal on se- lected areas in communication. These source titles provide essential works that pertain to AoI and UA V , respectively . Refer to T able 6 for more details about the titles and the per - centage of the document produced by these source titles. 4.5 Distribution of publications geographically The proportion of research contributions made by each of the topmost 18 countries is outlined in T able 7 . China con- tributes the most articles, with 64.75%, followed by the United States (US) (30.33%). This finding suggests that China is the main contributor to research concerning UA V and AoI, and the closest country is the USA. It is interest- ing to note that Hong Kong (12.30%), which is a special 98 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. T able 6: Most active source title with at least 3 publications Source title No of documents %(N=122) Citations %(N=622) IEEE internet of things journal 1 1 19.3 174 27.97 IEEE transactions on vehicular technology 9 15.79 181 29.1 IEEE transactions on wireless communication 6 10.53 83 13.34 IEEE transactions on intelligent transportation systems 5 8.77 38 6.1 1 IEEE vehicular technology conference 5 8.77 5 0.8 Proceedings- IEEE Infocom 4 7.02 34 5.47 IEEE international conference on communications 4 7.02 8 1.29 IEEE wireless communications and networking conferences (wcnc) 4 7.02 0 0 IEEE transactions on communications 3 5.26 55 8.84 IEEE journal on selected areas in communications 3 5.26 34 5.47 IEEE transactions on mobile computing 3 5.26 10 1.61 administrative region in China, is ranked third among the most productive countries. The fourth and fifth on the list are South Korea (7.38%) and Australia (6.56%), respec- tively . Countries such as Italy , Germany , and Singapore are located at the very bottom of the list. T ogether , they are responsible for less than 5% of all the publications. Apart from China and the United States, the study does not identify additional countries commonly prone to natural disasters where UA V applications are crucial [ 37 , 54 , 12 ]. Countries such as Indonesia, V ietnam, Malaysia, the Philip- pines, and T urkey are a few examples of this category . Re- searches centered on AoI and UA V are more frequently found in industrialized nations. In such regions, there are a lot of research sponsors that concentrate on UA V appli- cations to improve the ef ficiency of UA Vs through the AoI metric. T able 7: T op 18 countries/regions of the published articles Country/Region Frequency %(N=122) Citations China 79 64.75 653 United States 37 30.33 661 Hong kong 15 12.3 367 South Korea 9 7.38 92 Australia 8 6.56 54 Canada 7 5.74 97 Finland 6 4.92 23 Japan 5 4.1 23 Qatar 4 3.28 148 Sweden 4 3.28 31 United kingdom 3 2.46 82 Lebanon 3 2.46 70 Luxembour g 3 2.46 20 United Arab Emirates 3 2.46 10 India 3 2.46 7 Italy 3 2.46 1 Germany 2 1.64 43 Singapore 2 1.64 3 196 160.66 2385 4.6 Authorship The authors who have contributed the most to AoI-aware UA V deployment are listed in T able 8 . Liu J. (China) and Han Z. (United States) have the most documents, which puts the authors at the top of the table with 10 publications each. The third on the list is Zhang H. (United States) with 9 documents. Next is Poor H. V . (United States) and Zhang X. (China) in fourth and fifth with 8 documents each. W ang X. has 7 and Song I. has 6 documents. Five other authors have 5 documents each. The remaining authors were re- sponsible for at least four documents each. T able 8: Most productive authors (minimum document 4, citation 10) Author ’ s name Documents %(N=122) Citations Liu J. 10 8.2 244 Han Z. 10 8.2 86 Zhang H. 9 7.38 1 15 Poor H. V . 8 6.56 107 Zhang X. 8 6.56 18 W ang X. 7 5.74 218 Song I. 6 4.92 103 Bai B. 5 4.1 216 Fan P . 5 4.1 134 Letaief K. B. 5 4.1 134 Han R. 5 4.1 20 Y ang Y . 5 4.1 16 Dai H. 4 3.28 180 T ong P . 4 3.28 66 Qin X. 4 3.28 45 Chen X. 4 3.28 22 W ang W . 4 3.28 15 Shen C. 4 3.28 10 According to [ 55 ], collaboration between authors from a variety of disciplines is required to improve any sec- tor . Hence, an increase in international collaboration is re- quired. Fig. 5 and Fig. 6 illustrate the level of collabora- tion that exists between academics using authors and coun- tries as units of analysis and fractional counting methods. An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 99 Nonetheless, China, as well as the United States, are the two nations that are leading the char ge in the combined ef- forts (refer to Fig. 6 ). The analysis and visualization map demonstrates a comprehensive network of joint ef forts that spans all the continents. Chen X.’ s works, which involved a broader collaboration with researchers from diverse coun- tries, is the most prestigious of all of them (refer to Fig. 5 ). Although [ 56 ] stated that co-authorship preferences are determined and shaped by a variety of factors, including cultural relations, geopolitical position, and language, ac- cording to the findings of this study , geopolitical proxim- ity and shared language are two of the most important fac- tors that signify co-authorship relationships across coun- tries. Remarkably , there is a bigger quantity of research ar - ticles coming out from China as well as the USA. This could be a result of their interest in newer technological advance- ments. For example, the world-leading UA V manufacturer (DJI) is a Chinese technology company and has 76% global market share of consumer and commercial UA Vs. In addi- tion, [ 47 ] highlighted that academics working in the United States have a remarkable openness when it comes to work- ing with peers in other countries. 4.7 Most influential institutions The top institutions for AoI-aware UA V research are listed in T able 9 , which shows that the institutions have a t least 3 publications. According to the finding, each of the institu- tions has contributed at most 3 documents. The breakdown of the number of articles from each institution and their ci- tation counts has been demonstrated accordingly . Remark- ably , the Department of electronic engineering, T singhua University , Beijing China is the institution with the most ci- tation (131), which is followed by the Department of Elec- tronics, Peking University , Beijing China. 4.8 Citation analysis According to [ 57 ], the impact of particular research can be measured by the extent t o which other researchers have found it to be beneficial. The citation metrics of the 122 articles, which spans the years 2018 to 2023, are presented here in T able 10 . The total amount of citations throughout the course of the past 5 years is 1,073, which breaks down to 178.83 citations per year and 8.8 citations per paper . Ci- tations are supposed to illustrate the impact of a publica- tion in relation to several other publications according to the ideas of other researchers and their research findings. Con- sequently , the number of citations that are used in research evaluation serves as a determining factor of the impact of the research. Accordingly , the citations are used to indicate that a publication has improved the quality of several other publications [ 58 ]. Furthermore, T able 1 1 identifies the study by Liu et al. [ 59 ] as having the highest citation count. The highly refer - enced article, which is titled ”Age-optimal trajectory plan- ning for UA V -assisted data collection” is currently the most cited in the list with 1 17 citations. This study focuses on op- timizing the collection of data from ground sensor nodes using unmanned aerial vehicles (UA Vs) in wireless sen- sor networks. The goal is to plan the UA V’ s trajectory to minimize the age of information (AoI) gathered from the nodes. T wo types of optimal trajectories are considered: one aims to minimize the age of the oldest information col- lected, while the other aims to minimize the average age of information. The study shows that finding an optimal trajectory is equivalent to finding the shortest path in the sensor network. The dynamic programming method and genetic algorithm are used to find these trajectories. Sim- ulation results validate the ef fectiveness of the proposed methods and demonstrate how the UA V’ s trajectory is in- fluenced by the AoI metrics. Similarly , the second paper on the list is cited 1 13 [ 60 ], which has a slightly equal num- ber of citations with the first article on the list [ 59 ]. Both papers were published in 2018, and are technical papers. The study observed most of the papers are cited less than 50 times, except the study by [ 61 ] which is cited 78 times, respectively . 4.9 T extual analysis The VOSviewer is capable of recognizing and analyzing keywords, after which they will be presented in a structured format. A representation of a co-word network in the form of a map was developed using bibliographic information. It was possible to standardize the principles of involvement in relation to the keywords based on the strength of the con- nection between them [ 62 ]. T o graphically locate and place each word on the map, the approach known as ”visualiza- tion of similarities” was utilized accordingly [ 41 ]. In con- clusion, the VOSviewer method provides a variety of reso- lution parameters to enable the detection of a wide variety of clusters. In this study , the study focused and selected 21 key- words, which were used to measure the relative full strength of connections and co-occurrence with other keywords. The colors were applied to dif ferentiate between three unique groups (green, blue, and red). The graphical de- piction of co-words (keywords co-occurrence) is shown in Fig. 7 to Fig. 1 1 . Particularly , the network is produced with respect to the information contained in earlier litera- ture on AoI and UA V , as demonstrated in Fig. 7 . The in- vestigation of the terms is represented by clusters of vary- ing colors and sizes. The VOSviewer identified three dis- tinct clusters and assigned each one of those clusters one of three colors, based on the thematic community they were most closely associated. The size of the cluster represented by some terms shows the frequency of their occurrences in titles and abstracts of the publications [ 41 ]. Moreover , the observed distance between the cluster indicates the strength of their relationships. The number of times that both words appear together in the titles and abstracts of the various pa- pers provided evidence for this connection. The inclusion criteria of a term to be selected must have at least ten occur - 100 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. Figure 5: Map of the co-authorship based on Author unit of analysis Figure 6: Map of the co-authorship based on countries unit of analysis T able 9: Most influential institutions with a maximum of 3 publications Institution Frequency %(N=122) Citations Lab of rail traf fic control and safety , Beijing Jiaotong University , Beijing, China 3 2.46 3 Department of computer science and engineering, Kyung hee university , Seoul, South Korea 3 2.46 44 Department of electrical and computer engineering, university of Houston, Houston, T exas, united states 3 2.46 39 Department of electronic engineering, Beijing national research center for information science and technology , T singhua University , Beijing, China 3 2.46 131 Department of electronics, Peking University , Beijing, China 3 2.46 85 School of electrical engineering and computer science, Ningbo University , Zhejiang, China 3 2.46 65 School of electronics and communication engineering, Sun yat-sen university , Guangzhou, China 3 2.46 66 Laboratory of networking and switching technology , Beijing University of posts and telecommunications, Beijing, China 3 2.46 46 An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 101 T able 10: Citations metrics Measure Data Y ears of publications 2018—2023 Y ears of citations (6) 2018—2023 Quantity of papers 122 Citations count 1073 Citation/year 178.83 Citation/paper 8.8 Citations/author 3.17 Papers/author 0.36 Authors/paper 2.78 rences for both binary and full counting. In binary counting, the occurrences attribute reflects the number of documents in which a word appears at least once, whereas the occur - rences attribute in full counting displays the total number of appearances of a term across all documents [ 63 ]. Fig. 7 and Fig. 8 illustrate the co-occurrence network of terms from titles and abstracts fields based on full and binary counting, respectively , while Fig. 10 and Fig. 1 1 il- lustrate the co-occurrence network of terms from abstracts fields only; based on full and binary counting. According to the data, topics such as trajectory , data collection, AoI minimization, data freshness, average AoI, conver gence, and timeliness, are frequently used in the literature. More- over , the study has identified the usage of UA Vs in disrup- tive technologies such as wireless sensor network (WSN), mobile edge computing (MEC), and IoT , as well as, meth- ods such as the DRL, optimization algorithm, and Markov decision process (MDP) (see Fig. 7 . Secondly , the map of title and abstract based on full counting presents more keywords associated with UA V research. Example of these keywords includes status update, transmission, application, and cellular internet (see Fig. 8 ). Furthermore, ef fectiveness, communication, the data packet, and a few other keywords observed in Fig. 10 and Fig. 1 1 disclose the potential of UA V applications to im- prove overall data collection, accuracy , completeness, re- liability , relevance, and more importantly timeliness. Im- proving these characteristics will help management and decision-making processes in any UA V applications sce- nario. As a result, both the public and private sectors should concentrate their ef forts on developing their capabilities of UA Vs in order to improve their data management and decision-making procedures. Consequently , the classifica- tion of the articles that were carried out through the schema- tization of the subtitles and a brief explanation of the intent of those studies reveals that most of the prior studies stud- ied the concerns of the relation between UA V , information freshness, UA V trajectory , etc. 5 Discussion and matters arising Through the bibliometric analysis of the relevant literature, this study has presented some of the most pertinent terms re- lated to AoI in UA V -assisted wireless communication liter - ature. It is quite obvious that the success of ef ficiently guar - anteeing information freshness in wireless networks can- not be considered complete without a discussion of the role of UA Vs especially in disaster -stricken or hard-to-reach places. The UA V is increasingly required for the delivery of fresh data in several applications. As a result, this study il- lustrates the key themes identified by the network visualiza- tion presented in the preceding sections (refer to Fig. 7 and Fig. 8 ). This study believes that UA V -aided AoI minimiza- tion can transform the network applications where data are needed in real-time, which includes emer gencies and dis- asters, industrial IoT networks, etc. The analysis of terms relating to UA V and AoI shows dif ferent applications, con- cepts, and methods in the literature. One particular method, with a remarkable performance when compared with the existing benchmarks, and is evident across a number of ex- isting studies is DRL [ 64 , 65 , 66 ] with few examples pre- sented in T able 12 and 13 . DRL is a prominent machine learning-based method that facilitates autonomous control of the UA V and thus has been quite resourceful in UA V tra- jectory planning for AoI minimization. T able 12 presents the problems and focus areas addressed by DRL in UA V and AoI research. In particular , DRL has been applied to various aspects, such as ener gy harvesting, real-time data collection decisions, UA V altitude scheduling policies, and AoI optimization. In T able 13 , the various application of DRL methods in UA V and AoI literature are presented. This presents how existing scholars have utilized DRL to address some of the issues and problems pertinent to UA V - assisted AoI. 5.1 T echnology paradigm of DRL in the use of UA V for achieving minimal AoI The analysis of the visualization network has shown that the concept of UA V -aided information freshness has garnered significant attention in various technological paradigms, showcasing its versatility and wide applicability while ac- commodating dif ferent assistant technologies. One such paradigm is the utilization of reconfigurable intelligent sur - face (RIS), as explored by [ 74 ]. These surfaces, capable of manipulating the propagation of electromagnetic waves, of fer an innovative approach to enhance information fresh- ness in UA V systems. Additionally , the study of studying UA V -aided information freshness in IoT domain has wit- nessed substantial research ef forts. Several studies have in- vestigated the integration of UA Vs with IoT to improve in- formation freshness [ 66 , 7 , 9 , 3 , 4 , 5 , 6 ]. These works have explored various aspects to facilitate timely data collection, information transmission, and processing by leveraging the agility and mobility of UA Vs. In addition, WSN have also been an area of focus in the context of UA V -aided communication for achieving infor - mation freshness. The studies by [ 61 , 103 ] have inves- tigated the integration of UA Vs with WSNs to improve the freshness of information. By deploying UA Vs as mo- 102 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. T able 1 1: Most influential papers (highly cited; Min-20 citation) S/N Authors T itle Y ear Citation citation per year 1 Liu J., W ang X., Bai B., Dai H. Age-optimal trajectory planning for UA V -assisted data collection 2018 1 17 23.4 2 Abd-Elmagid M.A., Dhillon H.S. A verage peak age-of-information minimization in UA V -assisted IoT networks 2018 1 13 28.25 3 Abd-Elmagid M.A., Ferdowsi A., Dhillon H.S., Saad W . Deep reinforcement learning for minimizing age- of-information in UA V -assisted networks 2019 48 12 4 Jia Z., Qin X., W ang Z., Liu B. Age-based path planning and data acquisition in UA V -Assisted IoT networks 2019 29 7.25 5 T ong P ., Liu J., W ang X., Bai B., Dai H. UA V -Enabled age-optimal data collection in wire- less sensor networks 2019 29 7.25 6 Li W ., W ang L., Fei A. Minimizing Packet Expiration Loss with Path Planning in UA V -Assisted Data Sensing 2019 27 6.75 7 T ripathi V ., T alak R., Modiano E. Age Optimal Information Gathering and Dissemi- nation on Graphs 2019 26 6.5 8 Zhou C., He H., Y ang P ., L yu F ., W u W ., Cheng N., Shen X. Deep RL-based T rajectory Planning for AoI Min- imization in UA V -assisted IoT 2019 24 6 9 W an S., Lu J., Fan P ., Letaief K.B. T oward Big Data Processing in IoT : Path Planning and Resource Management of UA V Base Stations in Mobile-Edge Computing System 2020 46 15.33 10 Hu J., Zhang H., Song L., Schober R., Poor H.V . Cooperative Internet of UA Vs: Distributed T rajec- tory Design by Multi-Agent Deep Reinforcement Learning 2020 43 14.33 1 1 Y i M., W ang X., Liu J., Zhang Y ., Bai B. Deep reinforcement learning for fresh data collec- tion in UA V -assisted IoT networks 2020 36 12 12 Zhang S., Zhang H., Han Z., Poor H.V ., Song L. Age of Information in a Cellular Internet of UA Vs: Sensing and Communication T rade-Of f Design 2020 35 1 1.67 13 Samir M., Assi C., Sharafed- dine S., Ebrahimi D., Ghrayeb A. Age of Information A ware T rajectory Planning of UA Vs in Intelligent T ransportation Systems: A Deep Learning Approach 2020 32 10.67 14 Hu H., Xiong K., Qu G., Ni Q., Fan P ., Letaief K.B. AoI-Minimal T rajectory Planning and Data Col- lection in UA V -Assisted W ireless Powered IoT Networks 2021 78 39 15 Liu J., T ong P ., W ang X., Bai B., Dai H. UA V -Aided Data Collection for Information Freshness in W ireless Sensor Networks 2021 34 17 16 Samir M., Elhattab M., Assi C., Sharafeddine S., Ghrayeb A. Optimizing Age of Information through Aerial Re- configurable Intelligent Surfaces: A Deep Rein- forcement Learning Approach 2021 32 16 17 Abedin S.F ., Munir M.S., T ran N.H., Han Z., Hong C.S. Data Freshness and Ener gy-Ef ficient UA V Nav- igation Optimization: A Deep Reinforcement Learning Approach 2021 23 1 1.5 An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 103 Figure 7: Map of term co-occurrence network from title and abstract based on binary counting Figure 8: Map of term co-occurrence network from title and abstract based on full counting 104 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. T able 12: Problems and focus areas addressed by DRL in UA V and AoI research Classifications Focus areas References UA V ener gy Ener gy ef ficiency [ 67 , 68 , 69 ] Operation time [ 70 ] Ener gy consumption [ 71 ] Ener gy harvesting [ 65 ] Ener gy transfer [ 72 ] UA V trajectory T rajectory [ 70 , 71 , 73 ] UA V altitude [ 74 ] Distributed trajectory design [ 75 ] T rajectory design [ 76 ] T rajectory planning [ 69 , 77 , 78 , 79 , 68 ] T rajectory optimization [ 80 , 72 , 76 , 81 ] Path planning [ 82 , 83 , 84 , 68 ] Other focus UA V sensing [ 70 ] Sustainability [ 67 ] Sampling mode [ 85 ] Data collection [ 70 , 86 , 71 , 87 , 69 ] Surveillance Scheduling [ 71 , 74 , 88 , 73 ] T raining [ 71 ] Array signal processing [ 89 ] Conver gence [ 89 ] Resource management [ 89 ] Mobile relays [ 82 ] UA V altitude control [ 82 ] Unknown channel conditions [ 82 ] W ireless power transmission or transfer [ 90 , 91 , 92 ] Edge Caching [ 93 ] Mobile crowdsensing [ 94 ] Mobile data gathering centres [ 95 ] UA V -to-Device communication [ 96 , 97 ] T able 13: Aspects of DRL methods in UA V and AoI research Classifications Methods References DRL methods Multi-Agent Deep Reinforcement Learning (MADRL) [ 72 , 91 , 77 , 98 , 96 , 99 , 97 ] Neural combinatorial DRL [ 73 ] Deep Q-network (DQN) [ 89 ] Dec-POMDP [ 100 ] Graph convolutional reinforcement learning [ 94 ] Combined with DRL Heuristic algorithms [ 89 ] Optimization [ 89 ] Convex optimization [ 73 ] Federated learning [ 101 , 99 ] Stochastic games [ 98 ] Actor -critic algorithm [ 99 , 88 ] Scheduling Scheduling policy [ 82 ] Queuing policy [ 100 ] Of f-Policy; On-Policy [ 89 ] Proximal policy optimization (PPO) [ 89 , 66 , 74 ] An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 105 DRL-assisted W ireless Networks Reconfigurable intelligent surface (RIS) [ 89 , 87 , 74 ] V ehicular networks [ 88 ] UA V swarm [ 84 , 78 ] & Cooperative internet of UA Vs [ 75 , 71 ] IoT [ 85 , 71 , 89 , 68 , 92 , 74 , 79 , 32 ] & W ireless IoT networks [ 89 ] W ireless powered communication network (WPCN) [ 72 ] Heterogeneous network [ 84 ] Space-Air - Ground Integrated Net- work [ 102 ] Multi-access edge com- puting [ 98 ] Underwater linear net- works [ 95 ] Sensor Network [ 70 ] & W ireless sensor networks [ 70 ] Cellular (multi-cell & internet of UA Vs) network [ 70 , 96 , 97 ] Mobile edge computing (MEC) [ 101 , 82 , 93 , 83 , 99 , 78 ] Lar ge-scale wireless networks [ 89 ] Figure 9: Examples of AoI-aware UA V -aided networks architecture where DRL applications have been applied bile data collectors or relays, these studies have demon- strated the potential of UA V -WSN collaboration in achiev- ing real-time data updates and minimizing information stal- eness. Furthermore, MEC has emer ged as a promising paradigm for which UA V has been resourceful in enhanc- ing information freshness. For instance, the integration of UA V and MEC paradigms for of floading computation tasks and reducing latency , thereby ensuring fresh and up-to-date information was considered in [ 104 , 105 ]. These works highlight the potential of UA V -MEC collaboration in en- abling real-time data processing and analysis. Moreover , the realm of cellular networks has also witnessed research endeavours aiming to improve information freshness with the assistance of UA Vs. In [ 106 ], the integration of UA Vs into cellular networks was explored, leveraging their mobil- ity and flexibility to enhance the freshness of information. Fig. 9 captures the key technological paradigms in the context of DRL for UA V -aided information freshness. This provides a summary of the technological domains and their interconnections, illustrating the diverse applications of UA Vs in enhancing information freshness. Specifically , UA V -aided information freshness utilizing DRL has been extensively explored in various technological paradigms, including RIS, IoT , WSN, MEC, and cellular networks. The existing studies have demonstrated the potential of UA Vs in ensuring up-to-date and timely information up- dates, of fering valuable insights into the integration of UA Vs with dif ferent domains to enhance information fresh- ness in diverse applications. 5.2 Implications of the study This study conducted bibliometric analysis and provided knowledge on the major keyword, patterns of citations, 106 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. publication activities, and the state of cooperation among contributors throughout the course of the UA V -information freshness of existing research. Nevertheless, this study is not without limitations and thus provides suggestions for future studies. As a result, the study discussed the potential gap for future research, implications, and limitations. Firstly , this study highlights themes to motivate future re- search as a result of investigating bibliometric data of UA Vs for AoI minimization in a dif ferent range of scenarios. The examination of the bibliometric analysis showing the rela- tive priority of various elements illustrates the importance of particular subjects to the research community . This is extremely important at the moment because the notion of the AoI metric as a measure of information freshness for UA V -aided networks is been studied mainly within the past five years, and considering the fact the technology can be used in IoT , IIoT , RIS, WSN, etc. Therefore, it is necessary to emphasize the ef fectiveness of studying AoI to improve data freshness for UA V applications. In this regard, the bib- liometric analysis has shown to be useful because it visual- izes the network of key terms about the myriad aspects of research interest that could influence academic research in various fields. The bibliometric method, on the other hand, deviates from the way that earlier researchers reported literature studies. Few studies [ 107 , 108 , 109 , 73 ] reported literature and evaluated some of the contributions made about AoI- aided UA V . Although a brief literature report was used to show the literature gap and justify the objectives of the ex- isting studies in those works, the literature is not enough to present a holistic picture of the interest in AoI and UA V in various applications. In addition, narrative evaluation is both limited to a single issue of interest and may not capture a very wide scope in suf ficient detail. In addition, this study generates accurate, reliable, and enough bibliometric data for UA V and AoI. Earlier research utilized the VOSviewer for bibliometric analysis which em- phasizes the value of such in-depth text analysis and its findings [ 45 , 47 , 48 ]. Hence, the relevance of this research is in its ability to harmonize those aspects that are more rel- evant for UA V and AoI in real-time applications. Further - more, this research has other drawbacks. For example, the Scopus database is updated often, resulting in a fluctuating number of publications and citations [ 1 10 ]. Therefore, the accuracy of the data acquired from the Scopus database on a particular day is could be updated. Moreover , the co-word analysis (co-occurrence analysis of keywords) also has lim- its as certain publications may not be considered in biblio- metric records. Thus, the quality of the co-occurrence anal- ysis is dependent on the indexing method employed, over which this study has little control [ 1 1 1 ]. In light of this, it is proposed that future research utilize a unique method that blends qualitative and quantitative methods. The fun- damental constraint of this study is that the evaluation of 5-plus years of research in UA Vs and AoI is limited to ar - ticles published in associated publications. In addition, the choice of the keywords was also dependent on the assess- ment of the relevant literature and the definition of UA Vs and AoI; there may be other related keywords. This study is one of the few that analyzes the literature about UA V and AoI with bibliometric data. As a result, this adds more insight into the applications of UA V -aided AoI to enhance information freshness. Moreover , these findings demonstrate, in a nutshell, that researchers and academics should take action in order to enhance the development of algorithms and hardware that improves the use of UA V for various time-critical applications. The research commu- nity should be given enough support from research centres and universities to propagate the significance of AoI to im- prove information freshness. In a similar vein, suf ficient support in the form of enabling legislation and financial re- muneration is needed in order to improve interest in AoI, especially for practical, time-critical, real-time and emer - gency applications, in an ef ficient manner . Concerns about the information freshness should be used as a cornerstone to holistically improve techniques to improve the AoI per - formance of UA V -assisted wireless communication. More- over , the findings of this study may provide important in- formation and recommendations to policymakers, allowing them to further provide support to researchers and techni- cal experts in developing AoI-sensitive solutions for UA V - related applications. Secondly , the scientific mapping and profiling were based on quantitative approaches, which aid in the analysis of the reports and provide a full image of the study field, showing how significant the study themes are. Therefore, the conclusions of this study provide a sub- stantial body of evidence that can persuade researchers to take into consideration concerns that are essential for the implementation of UA V and AoI in real-time application scenarios. 5.3 Recommendations of the study Based on the comprehensive bibliometric analysis con- ducted in this study , several recommendations can be made to further advance research and practice in the field of unmanned aerial vehicles -assisted wireless communica- tion for information freshness. These recommendations are aimed at addressing important aspects that emer ged from the analysis and can guide future investigations and initia- tives in this area. These recommendations are discussed as follows; – Foster interdisciplinary collaboration: The findings of this study highlight the multidisciplinary nature of research on UA Vs and information freshness. T o fa- cilitate progress in this field, it is recommended to en- courage collaboration [ 55 ] among researchers from di- verse disciplines such as computer science, telecom- munications, transportation, and optimization. Inter - disciplinary teams can bring together complementary expertise and contribute to the development of innova- tive solutions for further facilitating AoI-aware UA V - aided interventions in various applications scenarios. An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 107 – Focus on further impr oving on DRL and optimiza- tion algorithms: The analysis revealed that deep reinforcement learning and optimization algorithms have been prominent methods applied in the context of UA Vs and information freshness. Future research should further explore and enhance these approaches to improve real-time status updates and ensure ef fi- cient information delivery . Investigating novel algo- rithms and techniques can contribute to the develop- ment of more robust and reliable UA V systems. – Addr ess energy efficiency and harvesting: Ener gy ef ficiency and harvesting emer ged as significant re- search topics in the context of UA Vs and information freshness. T o mitigate the limitations associated with limited onboard power and enhance the sustainabil- ity of UA V operations, it is recommended to investi- gate ener gy-ef ficient mechanisms and explore innova- tive approaches for ener gy harvesting or sources, such as solar , fuel cells, combustion engines, and kinetic ener gy , most of which can be utilized in UA V [ 22 ]. These ef forts can lead to longer flight durations and increased operational capabilities. – Enhance trajectory planning and design: The study identified trajectory planning and design as critical areas of interest for researchers. Future investiga- tions should focus on developing advanced algorithms and methodologies to optimize UA V trajectories, tak- ing into account factors such as communication con- straints, environmental conditions, and real-time data availability [ 23 , 18 ]. Ef fective trajectory planning can minimize delays, improve information freshness, and enhance overall UA V performance. – Emphasize scheduling strategies: The analysis high- lighted the significance of scheduling strategies in achieving information freshness (see T able 3 ). Re- searchers should explore scheduling techniques that consider time-critical applications and prioritize real- time information updates. Investigating dynamic scheduling algorithms and adaptive mechanisms can contribute to ef ficient resource allocation and ensure timely data delivery [ 64 , 66 ]. – Bridge the gap between r esear ch and practice: The implications derived from this study have practical significance. T o bridge the gap between research and practice, it is recommended to foster collaborations between academia, industry , and policymakers. In- dustry stakeholders can provide valuable insights into practical challenges and requirements, while policy- makers can facilitate the adoption and integration of UA V systems for information freshness in various do- mains. Similarly , practical UA V deployments on the field for evaluating some of the proposed algorithms would be very promising for advancing research in this area. By following these recommendations, researchers and practitioners can further advance UA V -aided solutions for information freshness, leading to the development of more ef ficient and reliable systems that cater to the demands of real-time status updates and time-critical applications. 6 Conclusion T echnology-driven various applications are continuously expanding as a new research topic. AoI in UA V -aided in- formation transmissions have attracted increasing interest, as an evolutionary paradigm. In view of this, there is an in- creasing drive to increase the ef fectiveness and ef ficiency of UA V communication and control to increase information freshness in several applications. This field of study is un- der going constant evolution. As result, each step towards improving methods, algorithms, procedures, or tools can make a substantial contribution to the literature. Therefore, the purpose of this study was to explore and analyze the bibliometric data of studies relating to the use of UA Vs for AoI-sensitive applications from the literature by using the VOSviewer bibliometric program. The use of bibliometric analysis revealed dif ferent primary and secondary aspects of the research that has been done in the field of UA V and AoI, via data analysis and visualizations. For instance, the most influential keywords, authors, universities, and coun- tries were all identified. Similarly , the text analysis using VOSviewer was used to determine the co-occurrence of the terms used by previous studies. After statistically an- alyzing the 122 articles, the study found that optimization, MDP and DRL as significant tools, while flight trajectory , scheduling, and ener gy ef ficiency are emer ging aspects of the current literature. In addition to that, the results of this study provide suggestions for enhancing research and prac- tical application deployment in this domain. Acknowledgments The authors thank the Centre for Intelligent Cloud Comput- ing (CICC) and the research management center Multime- dia University Malaysia for supporting this project. Refer ences [1] M. T ajik, M. Maleki, N. Mokari, M. Javan, H. Saeedi, B. Peng, and E. Jorswieck, “T wo-hop age of information scheduling for multi-uav as- sisted mobile edge,” 2022. Computing: FRL vs MADDPG. arXiv preprint arXiv:2206.09488., https://doi.or g/10.48550/arXiv .2206.09488, arxiv = 2206.09488. [2] B. Li, Z. Fei, Y . Zhang, and M. Guizani, “Secure uav communication networks over 5g,” IEEE W ir eless Communications , vol. 26, no. 5, p. 1 14–120, 2019. https://doi.or g/10.1 109/MWC.2019.1800458. 108 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. [3] S. Aggarwal, N. Kumar , and S. T anwar , “Blockchain-envisioned uav communication using 6g networks: Open issues, use cases, and future directions,” IEEE Internet of Things Journal , vol. 8, no. 7, p. 5416–5441, 2020. https://doi.or g/10.1 109/JIOT .2020.3020819. [4] J. Xu, Y . Zeng, and R. Zhang, “Uav-enabled wire- less power transfer: T rajectory design and ener gy optimization,” IEEE transactions on wir eless com- munications , vol. 17, no. 8, p. 5092–5106, 2018. https://doi.or g/10.1 109/TWC.2018.2838134. [5] H. W ang, X. Li, R. Jhaveri, T . Gadekallu, M. Zhu, T . Ahanger , and S. Khowaja, “Sparse bayesian learning based channel estimation in fbmc/oqam industrial iot networks,” Computer Communications , vol. 176, p. 40–45, 2021. https://doi.or g/10.1016/j.comcom.2021.05.020. [6] Y . Huang, W . Mei, J. Xu, L. Qiu, and R. Zhang, “Cognitive uav communication via joint maneuver and power control,” IEEE T ransactions on Com- munications , vol. 67, no. 1 1, p. 7872–7888, 2019. https://doi.or g/10.1 109/TCOMM.2019.2931322. [7] Q. Abbas, S. Zeb, and S. Hassan, “Age of infor - mation in backscatter communication,” in W ir eless- Power ed Backscatter Communications for Inter - net of Things , p. 67–80, Cham: Springer , 2021. https://doi.or g/10.1007/978-3-030-46201-7 5 . [8] A. Kosta, N. Pappas, and V . Angelakis, “Age of information: A new concept, met- ric, and tool,” Foundations and T r ends® in Networking , vol. 12, no. 3, p. 162–259, 2017. http://dx.doi.or g/10.1561/1300000060. [9] Q. Abbas, S. A. Hassan, H. K. Qureshi, K. Dev , and H. Jung, “A comprehensive survey on age of information in massive iot networks,” Com- puter Communications , vol. 197, p. 199–213, 2022. https://doi.or g/10.1016/j.comcom.2022.10.018. [10] J. Song, D. Gunduz, and W . Choi, “Optimal schedul- ing policy for minimizing age of information with a relay ,” 2020. arXiv preprint arXiv:2009.02716, https://doi.or g/10.48550/arXiv .2009.02716. [1 1] M. Shehzad, S. Hassan, M. Luque-Nieto, and P . Otero, “Uav trajectory optimization and choice for uav placement for data collection in be- yond 5g networks,” in Intelligent Unmanned Air V ehicles Communications for Public Safety Net- works , p. 133–144, Singapore: Springer , 2022. http://dx.doi.or g/10.1007/978-981-19-1292-4 6 . [12] X. Cui, Y . W ang, S. Y ang, H. Liu, and C. Mou, “Uav path planning method for data collection of fixed- point equipment in complex forest environment,” Fr ontiers in Neur or obotics , vol. 16, p. 1–16, 2022. https://doi.or g/10.3389/fnbot.2022.1 105177. [13] X. Li, J. Li, and D. Liu, “Ener gy-ef ficient uav trajectory design with information fresh- ness constraint via deep reinforcement learn- ing,” Mobile Information Systems , 2021. https://doi.or g/10.1 155/2021/1430512. [14] A. Arvanitaki and N. Pappas, “Modeling of a uav-based data collection system,” in 2017 IEEE 22nd International W orkshop on Computer Aided Modeling and Design of Communication Links and Networks (CAMAD , p. 1–6, IEEE, 2017. https://doi.or g/10.1 109/CAMAD.2017.8031526. [15] A. Khuwaja, Y . C hen, N. Zhao, M. Alouini, and P . Dobbins, “A survey of channel modeling for uav communications,” IEEE Communications Sur - veys T utorials , vol. 20, no. 4, p. 2804–2821, 2018. https://doi.or g/10.1 109/COMST .2018.2856587. [16] W . Khawaja, I. Guvenc, D. W . Matolak, U.-C. Fiebig, and N. Schneckenbur ger , “A survey of air -to- ground propagation channel modeling for unmanned aerial vehicles,” IEEE Communications Surveys & T utorials , vol. 21, no. 3, pp. 2361–2391, 2019. https://doi.or g/10.1 109/COMST .2019.2915069. [17] C. W ang, J. Huang, H. W ang, X. Gao, X. Y ou, and Y . Hao, “6g wireless channel measurements and models: T rends and challenges,” IEEE V ehicular T echnology Magazine , vol. 15, no. 4, p. 22–32, 2020. https://doi.or g/10.1 109/MVT .2020.3018436. [18] O. A. Amodu, U. A. Bukar , R. A. R. Mahmood, C. Jarray , and M. Othman, “Age of information min- imization in uav-aided data collection for wsn and iot applications: A systematic review ,” Journal of Net- work and Computer Applications , p. 103652, 2023. https://doi.or g/10.1016/j.jnca.2023.103652. [19] Z. Qadir , F . Ullah, H. Munawar , and F . Al- T urjman, “Addressing disasters in smart cities through uavs path planning and 5g commu- nications: A systematic review ,” Computer Communications , vol. 168, p. 1 14–135, 2021. https://doi.or g/10.1016/j.comcom.2021.01.003. [20] W . P . Coutinho, M. Battarra, and J. Fliege, “The un- manned aerial vehicle routing and trajectory optimi- sation problem, a taxonomic review ,” Computers & Industrial Engineering , vol. 120, pp. 1 16–128, 2018. https://doi.or g/10.1016/j.cie.2018.04.037. [21] N. Islam, M. M. Rashid, F . Pasandideh, B. Ray , S. Moore, and R. Kadel, “A review of appli- cations and communication technologies for in- ternet of things (iot) and unmanned aerial ve- hicle (uav) based sustainable smart farming,” Sustainability , vol. 13, no. 4, p. 1821, 2021. https://doi.or g/10.3390/su13041821. [22] A. T ownsend, I. N. Jiya, C. Martinson, D. Bessarabov , and R. Gouws, “A comprehensive review of ener gy sources for unmanned aerial vehi- cles, their shortfalls and opportunities for improve- ments,” Heliyon , vol. 6, no. 1 1, p. e05285, 2020. https://doi.or g/10.1016/j.heliyon.2020.e05285. [23] O. A. Amodu, R. Nordin, C. Jarray , U. A. Bukar , R. A. Raja Mahmood, and M. Othman, “A survey on the design aspects and opportu- An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 109 nities in age-aware uav-aided data collection for sensor networks and internet of things applica- tions,” Dr ones , vol. 7, no. 4, p. 260, 2023. https://doi.or g/10.3390/drones7040260. [24] F . Gambella, L. Sistu, D. Piccirilli, S. Cor - posanto, M. Caria, E. Arcangeletti, A. R. Proto, G. Chessa, and A. Pazzona, “Forest and uav: a bibliometric review ,” Contemporary Engineering Sciences , vol. 9, no. 28, pp. 1359–1370, 2016. http://dx.doi.or g/10.12988/ces.2016.68130. [25] E. Raparelli and S. Bajocco, “A bibliometric analysis on the use of unmanned aerial vehicles in agricultural and forestry studies,” International Journal of Re- mote Sensing , vol. 40, no. 24, pp. 9070–9083, 2019. https://doi.or g/10.1080/01431 161.2019.1569793. [26] A. P . Singh, A. Y erudkar , V . Mariani, L. Iannelli, and L. Glielmo, “A bibliometric review of the use of unmanned aerial vehicles in precision agricul- ture and precision viticulture for sensing applica- tions,” Remote Sensing , vol. 14, no. 7, p. 1604, 2022. https://doi.or g/10.3390/rs14071604. [27] J. W ang, S. W ang, D. Zou, H. Chen, R. Zhong, H. Li, W . Zhou, and K. Y an, “Social network and bibliometric analysis of unmanned aerial vehicle remote sensing applications from 2010 to 2021,” Remote Sensing , vol. 13, no. 15, p. 2912, 2021. https://doi.or g/10.3390/rs13152912. [28] Y . Jiang, Y . Gao, W . Song, Y . Li, and Q. Quan, “Bibliometric analysis of uav swarms,” Jour - nal of Systems Engineering and Electr on- ics , vol. 33, no. 2, pp. 406–425, 2022. https://doi.or g/10.23919/JSEE.2022.000042. [29] S. Gokool, M. Mahomed, R. Kunz, A. Clulow , M. Sibanda, V . Naiken, K. Chetty , and T . Mab- haudhi, “Crop monitoring in smallholder farms us- ing unmanned aerial vehicles to facilitate precision agriculture practices: a scoping review and bib- liometric analysis,” Sustainability , vol. 15, no. 4, p. 3557, 2023. https://doi.or g/10.3390/su15043557. [30] S. K. Chauhan, “Scholarly output on drone research: A bibliometric study ,” DESIDOC Journal of Library & Information T echnology , vol. 39, no. 2, 2019. https://doi.or g/10.14429/DJLIT .39.2.13970. [31] M. V ideras Rodríguez, S. G. Melgar , A. S. Cordero, and J. M. A. Márquez, “A critical review of un- manned aerial vehicles (uavs) use in architecture and urbanism: Scientometric and bibliometric analysis,” Applied Sciences , vol. 1 1, no. 21, p. 9966, 2021. https://doi.or g/10.3390/app1 1219966. [32] A. I. Lawal, O. J. Ojo, M. Kim, and S. Kwon, “De- termination of blast-induced flyrock using a drone technology: a bibliometric overview with practi- cal soft computing implementation,” Arabian Jour - nal of Geosciences , vol. 15, no. 19, p. 1581, 2022. http://dx.doi.or g/10.1007/s12517-022-10770-7. [33] H. d. S. Felipetto, E. Mercante, O. H. V iana, D. C. Granemann, and A. R. Elias, “Uav applica- tions in wheat crop: a bibliometric approach to the literature,” Revista Ciência Agr onômica , vol. 54, p. e20228592, 2023. https://doi.or g/10.5935/1806- 6690.20230025. [34] M. Galina, M. Asvial, and M. Suryanegara, “A bibliometric analysis of unmanned aerial vehi- cles (uav) implementation in cellular network,” in 2022 1 1th Electrical Power , Electr onics, Communications, Contr ols and Informatics Seminar (EECCIS) , pp. 254–259, IEEE, 2022. https://doi.or g/10.1 109/EECCIS54468.2022.9902960. [35] G.-G. Hognogi, A.-M. POP , and A.-C. Marian- Potra, “Using uav for the digitalisation of public ad- ministration: A bibiometric analysis,” Studia Uni- versitatis Babes-Bolyai, Geographia , vol. 67, 2022. http://dx.doi.or g/10.24193/subbgeogr .2022.01. [36] B. Madani and M. Ndiaye, “Optimization of hybrid truck-drone delivery systems: A bib- liometric analysis,” in 2022 IEEE International Confer ence on T echnology Management, Opera- tions and Decisions (ICTMOD) , pp. 1–6, IEEE, 2022. https://doi.or g/10.1 109/ICTMOD55867.2022. 10041873. [37] E. Blasch, “Dddas advantages from high- dimensional simulation,” in 2018 W inter Simulation Confer ence (WSC , p. 1418–1429, IEEE, 2018. https://doi.or g/10.1 109/WSC.2018.8632336. [38] L. La, F . Xu, and D. Buhalis, “Knowl- edge mapping of sharing accommodation: A bibliometric analysis,” T ourism manage- ment perspectives , vol. 40, p. 100897, 2021. https://doi.or g/10.1016/j.tmp.2021.100897. [39] S. Kozhakhmet, Y . Rofcanin, A. Nur gabdeshov , and M. Las Heras, “A bibliometric analysis of psychological contract research: current status development and future research direc- tions,” International Journal of Manpower , 2023. https://doi.or g/10.1 108/IJM-01-2021-0009. [40] H. Gao and X. Ding, “The research landscape on the artificial intelligence: a bibliometric analysis of recent 20 years,” Multimedia T ools and Appli- cations , vol. 81, no. 9, pp. 12973–13001, 2022. http://dx.doi.or g/10.1007/s1 1042-022-12208-4. [41] N. Eck and L. W altman, “V isualizing biblio- metric networks,” in Measuring scholarly impact , p. 285–320, Cham: Springer , 2014. https://doi.or g/10.1007/978-3-319-10377-8 1 3. [42] E. Garfield and A. W elljams-Dorof, “Citation data: their use as quantitative indicators for science and technology evaluation and policy-making,” Science and Public Policy , vol. 19, no. 5, p. 321–327, 1992. https://doi.or g/10.1093/spp/19.5.321. [43] E. Garfield, “Historiographic mapping of knowl- edge domains literature,” Journal of Informa- tion Science , vol. 30, no. 2, p. 1 19–145, 2004. https://doi.or g/10.1 177/0165551504042802. [44] T . Nisonger , “A methodological issue concerning the 1 10 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. use of social sciences citation index journal citation reports impact factor data for journal ranking,” Li- brary Acquisitions: Practice Theory , vol. 18, no. 4, p. 447–458, 1994. 10.1016/0364-6408(94)90052-3. [45] S. Øyna and I. Alon, “A review of born glob- als,” International Studies of Management Or - ganization , vol. 48, no. 2, p. 157–180, 2018. https://doi.or g/10.1080/00208825.2018.1443737. [46] C. Bor gman, “Data citation as a bibliomet- ric oxymoron,” Theories of informetrics and scholarly communication , p. 93–1 15, 2016. https://doi.or g/10.1515/97831 10308464-008. [47] A. V aio, R. Hassan, and C. Alavoine, “Data intelligence and analytics: A bibliomet- ric analysis of human–artificial intelligence in public sector decision-making ef fective- ness,” T echnological For ecasting and So- cial Change , vol. 174, p. 121201, 2022. https://doi.or g/10.1016/j.techfore.2021.121201. [48] Z. Xu, Z. Ge, X. W ang, and M. Skare, “Bibliomet- ric analysis of technology adoption literature pub- lished from 1997 to 2020,” T echnological For ecast- ing and Social Change , vol. 170, p. 120896, 2021. https://doi.or g/10.1016/j.techfore.2021.120896. [49] A. Fink, Conducting r esear ch literatur e r eviews: Fr om the internet to paper . Sage publications, 2019. ISBN: 9781544318479. [50] J. Hirsch, “An index to quantify an individ- ual’ s scientific research output that takes into ac- count the ef fect of multiple coauthorship,” Sci- entometrics , vol. 85, no. 3, p. 741–754, 2010. http://dx.doi.or g/10.1007/s1 1 192-010-0193-9. [51] G. Marzi, M. Dabić, T . Daim, and E. Garces, “Product and process innovation in manufac- turing firms: a 30-year bibliometric analysis,” Scientometrics , vol. 1 13, pp. 673–704, 2017. http://dx.doi.or g/10.1007/s1 1 192-017-2500-1. [52] R. Rialti, G. Marzi, C. Ciappei, and D. Busso, “Big data and dynamic capabilities: a bibliometric analysis and systematic literature review ,” Manage- ment Decision , vol. 57, no. 8, pp. 2052–2068, 2019. https://doi.or g/10.1 108/MD-07-2018-0821. [53] D. Y u, Z. Xu, and W . W ang, “A bibliomet- ric analysis of fuzzy optimization and decision making (2002–2017),” Fuzzy Optimization and Decision Making , vol. 18, pp. 371–397, 2019. https://link.springer .com/article/10.1007/s10700- 018-9301-8. [54] E. Blasch, D. Bernstein, and M. Rangaswamy , “In- troduction to dynamic data driven applications sys- tems,” in Handbook of Dynamic Data Driven Ap- plications Systems , p. 1–25, Cham: Springer , 2018. http://dx.doi.or g/10.1007/978-3-319-95504-9. [55] J. T urner and R. Baker , “collaborative re- search: techniques for conducting collabora- tive research from the science of team sci- ence (scits),” Advances in Developing Human Resour ces , vol. 22, no. 1, p. 72–86, 2020. https://doi.or g/10.1 177/1523422319886300. [56] A. Schubert and G. Schubert, “Interna- tionality at university level,” Scientomet- rics , vol. 123, no. 3, p. 1341–1364, 2020. http://dx.doi.or g/10.1007/s1 1 192-020-03443-3. [57] L. Bornmann, R. Mutz, C. Neuhaus, and H. Daniel, “Citation counts for research evaluation: standards of good practice for analyzing bibliometric data and presenting and interpreting results,” Ethics in sci- ence and envir onmental politics , vol. 8, no. 1, p. 93– 102, 2008. https://doi.or g/10.3354/esep00084. [58] L. Bornmann and H. Daniel, “What do we know about the h index?,” Journal of the Amer - ican Society for Information Science and tech- nology , vol. 58, no. 9, p. 1381–1385, 2007. https://doi.or g/10.1002/asi.20609. [59] J. Liu, X. W ang, B. Bai, and H. Dai, “Age- optimal trajectory planning for uav-assisted data collection,” in IEEE INFOCOM 2018-IEEE Con- fer ence on Computer Communications W orkshops (INFOCOM WKSHPS , p. 553–558, IEEE, 2018. https://doi.or g/10.1 109/INFCOMW .2018.8406973. [60] M. Abd-Elmagid and H. Dhillon, “A verage peak age-of-information minimization in uav-assisted iot networks,” IEEE T ransactions on V ehicular T ech- nology , vol. 68, no. 2, p. 2003–2008, 2018. https://doi.or g/10.1 109/TVT .2018.2885871. [61] J. Liu, P . T ong, X. W ang, B. Bai, and H. Dai, “Uav-aided data collection for informa- tion freshness in wireless sensor networks,” IEEE T ransactions on W ir eless Communica- tions , vol. 20, no. 4, pp. 2368–2382, 2020. https://doi.or g/10.1 109/TWC.2020.3041750. [62] N. Eck and L. W altman, “V os: A new method for visualizing similarities between objects,” in Ad- vances in data analysis , p. 299–306, Berlin, Heidel- ber g: Springer , 2007. http://dx.doi.or g/10.1007/978- 3-540-70981-7 3 4. [63] N. Eck and L. W altman, “Manual for vosviewer ver - sion 1.6.17,” CWTS Meaningful Metrics. Univer - siteit Leiden , 2021. [64] M. A. Abd-Elmagid, A. Ferdowsi, H. S. Dhillon, and W . Saad, “Deep reinforcement learning for minimizing age-of-information in uav-assisted networks,” in 2019 IEEE Global Communica- tions Confer ence (GLOBECOM) , pp. 1–6, IEEE, 2019. https://doi.or g/10.1 109/GLOBECOM38437. 2019.9013924. [65] N. Zhang, J. Liu, L. Xie, and P . T ong, “A deep re- inforcement learning approach to ener gy-harvesting uav-aided data collection,” in 2020 International Confer ence on W ir eless Communications and Signal Pr ocessing (WCSP) , pp. 93–98, IEEE, 2020. https://doi.or g/10.1 109/WCSP49889.2020.9299806. [66] M. Samir , C. Assi, S. Sharafeddine, and A. Ghrayeb, “Online altitude control and scheduling policy An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 1 1 1 for minimizing aoi in uav-assisted iot wireless networks,” IEEE T ransactions on Mobile Com- puting , vol. 21, no. 7, pp. 2493–2505, 2020. https://doi.or g/10.1 109/TMC.2020.3042925. [67] E. Eldeeb, J. M. de Souza Sant’Ana, D. E. Pérez, M. Shehab, N. H. Mahmood, and H. Alves, “Multi-uav path learning for age and power opti- mization in iot with uav battery rechar ge,” IEEE T ransactions on V ehicular T echnology , 2022. https://doi.or g/10.1 109/TVT .2022.3222092. [68] E. Eldeeb, D. E. Pérez, J. M. de Souza Sant’Ana, M. Shehab, N. H. Mahmood, H. Alves, and M. Latva-Aho, “A learning-based trajectory plan- ning of multiple uavs for aoi minimization in iot networks,” in 2022 Joint Eur opean Confer ence on Networks and Communications & 6G Summit (EuCNC/6G Summit) , pp. 172–177, IEEE, 2022. https://doi.or g/10.1 109/EuCNC/6GSummit54941. 2022.9815722. [69] M. Sun, X. Xu, X. Qin, and P . Zhang, “Aoi- ener gy-aware uav-assisted data collection for iot networks: A deep reinforcement learn- ing method,” IEEE Internet of Things Jour - nal , vol. 8, no. 24, pp. 17275–17289, 2021. https://doi.or g/10.1 109/JIOT .2021.3078701. [70] C. Zhan, H. Hu, J. W ang, Z. Liu, and S. Mao, “T radeof f between age of information and opera- tion time for uav sensing over multi-cell cellular net- works,” IEEE T ransactions on Mobile Computing , 2023. https://doi.or g/10.1 109/TMC.2023.3267656. [71] X. W ang, M. Y i, J. Liu, Y . Zhang, M. W ang, and B. Bai, “Cooperative data collec- tion with multiple uavs for information freshness in the internet of things,” IEEE T ransactions on Communications , 2023. https://doi.or g/10.1 109/TCOMM.2023.3255240. [72] O. S. Oubbati, M. Atiquzzaman, H. Lim, A. Rachedi, and A. Lakas, “Synchronizing uav teams for timely data collection and ener gy transfer by deep rein- forcement learning,” IEEE T ransactions on V ehicu- lar T echnology , vol. 71, no. 6, pp. 6682–6697, 2022. https://doi.or g/10.1 109/TVT .2022.3165227. [73] A. Ferdowsi, M. A. Abd-Elmagid, W . Saad, and H. S. Dhillon, “Neural combinatorial deep rein- forcement learning for age-optimal joint trajec- tory and scheduling design in uav-assisted net- works,” IEEE Journal on Selected Ar eas in Com- munications , vol. 39, no. 5, pp. 1250–1265, 2021. https://doi.or g/10.1 109/JSAC.2021.3065049. [74] M. Samir , M. Elhattab, C. Assi, S. Sharafed- dine, and A. Ghrayeb, “Optimizing age of in- formation through aerial reconfigurable intelli- gent surfaces: A deep reinforcement learning ap- proach,” IEEE T ransactions on V ehicular T ech- nology , vol. 70, no. 4, pp. 3978–3983, 2021. https://doi.or g/10.1 109/TVT .2021.3063953. [75] J. Hu, H. Zhang, L. Song, R. Schober , and H. V . Poor , “Cooperative internet of uavs: Distributed trajectory design by multi-agent deep reinforce- ment learning,” IEEE T ransactions on Communi- cations , vol. 68, no. 1 1, pp. 6807–6821, 2020. https://doi.or g/10.1 109/TCOMM.2020.3013599. [76] X. Li, B. Y in, J. Y an, X. Zhang, and R. W ei, “Joint power control and uav trajectory design for information freshness via deep reinforcement learning,” in 2022 IEEE 95th V ehicular T ech- nology Confer ence:(VTC2022-Spring) , pp. 1–5, IEEE, 2022. https://doi.or g/10.1 109/VTC2022- Spring54318.2022.9860365. [77] K. Chi, F . Li, F . Zhang, M. W u, and C. Xu, “Aoi optimal trajectory planning for cooperative uavs: A multi-agent deep reinforcement learning ap- proach,” in 2022 IEEE 5th International Confer ence on Electr onic Information and Communica- tion T echnology (ICEICT) , pp. 57–62, IEEE, 2022. https://doi.or g/10.1 109/ICEICT55736.2022.9909005. [78] W . Fan, K. Luo, S. Y u, Z. Zhou, and X. Chen, “Aoi-driven fresh situation awareness by uav swarm: Collaborative drl-based ener gy-ef ficient trajectory control and data processing,” in 2020 IEEE/CIC International Confer ence on Communi- cations in China (ICCC) , pp. 841–846, IEEE, 2020. https://doi.or g/10.1 109/ICCC49849.2020.9238897. [79] C. Zhou, H. He, P . Y ang, F . L yu, W . W u, N. Cheng, and X. Shen, “Deep rl-based trajectory planning for aoi minimization in uav-assisted iot,” in 2019 1 1th International Confer ence on W ir eless Communica- tions and Signal Pr o cessing (WCSP) , pp. 1–6, IEEE, 2019. https://doi.or g/10.1 109/WCSP .2019.8928091. [80] H. Hu, K. Xiong, G. Qu, Q. Ni, P . Fan, and K. Letaief, “Aoi-minimal trajectory plan- ning and data collection in uav-assisted wireless powered iot networks,” IEEE Internet of Things Journal , vol. 8, no. 2, p. 121 1–1223, 2020. https://doi.or g/10.1 109/JIOT .2020.3012835. [81] S. F . Abedin, M. S. Munir , N. H. T ran, Z. Han, and C. S. Hong, “Data freshness and ener gy- ef ficient uav navigation optimization: A deep reinforcement learning approach,” IEEE T rans- actions on Intelligent T ransportation Sys- tems , vol. 22, no. 9, pp. 5994–6006, 2020. https://doi.or g/10.1 109/TITS.2020.3039617. [82] Y . Peng, Y . Liu, D. Li, and H. Zhang, “Deep rein- forcement learning based freshness-aware path plan- ning for uav-assisted edge computing networks with device mobility ,” Remote Sensing , vol. 14, no. 16, p. 4016, 2022. https://doi.or g/10.3390/rs14164016. [83] H. Chen, X. Qin, Y . Li, and N. Ma, “Ener gy- aware path planning for obtaining fresh updates in uav-iot mec systems,” in 2022 IEEE W ir e- less Communications and Networking Confer - ence (WCNC) , pp. 1791–1796, IEEE, 2022. https://doi.or g/10.1 109/WCNC51071.2022.9771867. [84] L. Shi, X. Zhang, X. Xiang, Y . Zhou, and 1 12 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. S. Sun, “Age of information optimization with heterogeneous uavs based on deep reinforce- ment learning,” in 2022 14th International Confer ence on Advanced Computational In- telligence (ICACI , p. 239–245, IEEE, 2022. https://doi.or g/10.1 109/ICACI55529.2022.9837720. [85] Z. Li, P . T ong, J. Liu, X. W ang, L. Xie, and H. Dai, “Learning-based data gathering for information freshness in uav-assisted iot net- works,” IEEE Internet of Things Journal , 2022. https://doi.or g/10.1 109/JIOT .2022.3215521. [86] Y . Hu, Y . Liu, A. Kaushik, C. Masouros, and J. Thompson, “T imely data collection for uav- based iot networks: A deep reinforcement learn- ing approach,” IEEE Sensors Journal , 2023. https://doi.or g/10.1 109/JSEN.2023.3265935. [87] X. Fan, M. Liu, Y . Chen, S. Sun, Z. Li, and X. Guo, “Ris-assisted uav for fresh data collection in 3d urban environments: A deep reinforcement learning approach,” IEEE T ransactions on V ehicu- lar T echnology , vol. 72, no. 1, pp. 632–647, 2022. https://doi.or g/10.1 109/TVT .2022.3203008. [88] M. Samir , C. Assi, S. Sharafeddine, D. Ebrahimi, and A. Ghrayeb, “Age of information aware trajectory planning of uavs in intelligent trans- portation systems: A deep learning approach,” IEEE T ransactions on V ehicular T echnol- ogy , vol. 69, no. 1 1, pp. 12382–12395, 2020. https://doi.or g/10.1 109/TVT .2020.3023861. [89] M. Sherman, S. Shao, X. Sun, and J. Zheng, “Opti- mizing aoi in uav-ris assisted iot networks: Of f pol- icy vs. on policy ,” IEEE Internet of Things Journal , 2023. https://doi.or g/10.1 109/JIOT .2023.3246925. [90] L. Liu, K. Xiong, J. Cao, Y . Lu, P . Fan, and K. B. Letaief, “A verage aoi minimization in uav-assisted data collection with rf wireless power transfer: A deep reinforcement learning scheme,” IEEE Inter - net of Things Journal , vol. 9, no. 7, pp. 5216–5228, 2021. https://doi.or g/10.1 109/JIOT .2021.31 10138. [91] Z. Li, J. Liu, L. Xie, X. W ang, and M. Jin, “A deep reinforcement learning approach for multi-uav-assisted data collection in wireless pow- ered iot networks,” in 2022 14th International Confer ence on W ir eless Communications and Signal Pr ocessing (WCSP) , pp. 44–49, IEEE, 2022. https://doi.or g/10.1 109/WCSP55476.2022.10039294. [92] Q. Dang, Q. Cui, Z. Gong, X. Zhang, X. Huang, and X. T ao, “Aoi oriented uav trajectory plan- ning in wireless powered iot networks,” in 2022 IEEE W ir eless Communications and Networking Confer ence (WCNC) , pp. 884–889, IEEE, 2022. https://doi.or g/10.1 109/WCNC51071.2022.9771588. [93] N. Hu, X. Qin, N. Ma, Y . Liu, Y . Y ao, and P . Zhang, “Ener gy-ef ficient caching and task of floading for timely status updates in uav- assisted vanets,” in 2022 IEEE/CIC Inter - national Confer ence on Communications in China (ICCC) , pp. 1032–1037, IEEE, 2022. https://doi.or g/10.1 109/ICCC55456.2022.9880683. [94] Z. Dai, C. Liu, Y . Y e, R. Han, Y . Y uan, G. W ang, and J. T ang, “Aoi-minimal uav crowdsensing by model-based graph convo- lutional reinforcement learning,” in IEEE IN- FOCOM 2022-IEEE Confer ence on Computer Communications , p. 1029–1038, IEEE, 2022. https://doi.or g/10.1 109/INFOCOM48880.2022. 9796732. [95] A. A. Al-Habob, O. A. Dobre, and H. V . Poor , “Age-optimal information gathering in linear un- derwater networks: A deep reinforcement learning approach,” IEEE T ransactions on V ehicular T ech- nology , vol. 70, no. 12, pp. 13129–13138, 2021. https://doi.or g/10.1 109/TVT .2021.31 17536. [96] F . W u, H. Zhang, J. W u, Z. Han, H. V . Poor , and L. Song, “Uav-to-device underlay communications: Age of information mini- mization by multi-agent deep reinforcement learning,” IEEE T ransactions on Communi- cations , vol. 69, no. 7, pp. 4461–4475, 2021. https://doi.or g/10.1 109/TCOMM.2021.3065135. [97] F . W u, H. Zhang, J. W u, L. Song, Z. Han, and H. V . Poor , “Aoi minimization for uav-to-device underlay communication by multi-agent deep reinforcement learning,” in GLOBECOM 2020-2020 IEEE Global Communications Confer ence , pp. 1–6, IEEE, 2020. https://doi.or g/10.1 109/GLOBECOM42002.2020. 9322539. [98] X. Chen, C. W u, T . Chen, Z. Liu, H. Zhang, M. Ben- nis, H. Liu, and Y . Ji, “Information freshness-aware task of floading in air -ground integrated edge com- puting systems,” IEEE Journal on Selected Ar eas in Communications , vol. 40, no. 1, pp. 243–258, 2021. https://doi.or g/10.1 109/JSAC.2021.3126075. [99] Z. Zhu, S. W an, P . Fan, and K. B. Letaief, “An edge federated marl approach for timeliness main- tenance in mec collaboration,” in 2021 IEEE International Confer ence on Communications W orkshops (ICC W orkshops) , pp. 1–6, IEEE, 2021. https://doi.or g/10.1 109/ICCW orkshops50388.2021. 9473729. [100] J. Xu, X. Jia, and Z. Hao, “Research on infor - mation freshness of uav-assisted iot networks based on ddqn,” in 2022 IEEE 5th Advanced Information Management, Communicates, Electr onic and Automation Contr ol Confer - ence (IMCEC) , vol. 5, pp. 427–433, IEEE, 2022. https://doi.or g/10.1 109/IMCEC55388.2022.10019931. [101] W . Fan, Q. W ei, Z. Zhou, S. Y u, and X. Chen, “A research on collaborative uavs intelligent decision optimization for aoi-driven federated learning,” Journal of Electr onics Information T echnology ( JEIT) , vol. 44, pp. 1–10, 2022. http://dx.doi.or g/10.1 1999/JEIT21 1406. [102] X. W ei, G. Zhang, and Z. Han, “Satellite-controlled An Exploratory Bibliometric Analysis of the Literature… Informatica 47 (2023) 91–1 14 1 13 uav-assisted iot information collection with deep reinforcement learning and device match- ing,” in 2022 7th International Confer ence on Intelligent Computing and Signal Pr o- cessing (ICSP) , pp. 1254–1259, IEEE, 2022. https://doi.or g/10.1 109/ICSP54964.2022.9778637. [103] P . T ong, J. Liu, X. W ang, B. Bai, and H. Dai, “Uav-enabled age-optimal data collection in wireless sensor networks,” in 2019 IEEE In- ternational Confer ence on Communications W orkshops (ICC W orkshops) , pp. 1–6, IEEE, 2019. https://doi.or g/10.1 109/ICCW .2019.8756665. [104] B. Liu, W . Zhang, W . Chen, H. Huang, and S. Guo, “Online computation of floading and traf- fic routing for uav swarms in edge-cloud com- puting,” IEEE T ransactions on V ehicular T ech- nology , vol. 69, no. 8, p. 8777–8791, 2020. https://doi.or g/10.1 109/TVT .2020.2994541. [105] S. W an, J. Lu, P . Fan, and K. B. Letaief, “T o- ward big data processing in iot: Path planning and resource management of uav base stations in mobile-edge computing system,” IEEE Internet of Things Journal , vol. 7, no. 7, pp. 5995–6009, 2019. https://doi.or g/10.1 109/JIOT .2019.2954825. [106] S. Zhang, H. Zhang, Z. Han, H. Poor , and L. Song, “Age of information in a cellular inter - net of uavs: Sensing and communication trade- of f design,” IEEE T ransactions on W ir eless Com- munications , vol. 19, no. 10, p. 6578–6592, 2020. https://doi.or g/10.1 109/TWC.2020.3004162. [107] M. Abd-Elmagid, N. Pappas, and H. Dhillon, “On the role of age of information in the internet of things,” IEEE Communications Magazine , vol. 57, no. 12, p. 72–77, 2019. https://doi.or g/10.1 109/MCOM.001.1900041. [108] M. Y i, X. W ang, J. Liu, Y . Zhang, and B. Bai, “Deep reinforcement learning for fresh data collection in uav-assisted iot networks,” in IEEE INFOCOM 2020-IEEE Confer ence on Computer Communications W orkshops (IN- FOCOM W orkshops) , pp. 716–721, IEEE, 2020. https://doi.or g/10.1 109/INFOCOMWKSHPS50562. 2020.9162896. [109] G. Ahani, D. Y uan, and Y . Zhao, “Age-optimal uav scheduling for data collection with bat- tery rechar ging,” IEEE Communications Let- ters , vol. 25, no. 4, pp. 1254–1258, 2020. https://doi.or g/10.1 109/LCOMM.2020.3047909. [1 10] L. V alenzuela-Fernandez, J. Merigó, J. Lichten- thal, and C. Nicolas, “A bibliometric analysis of the first 25 years of the journal of business- to-business marketing,” Journal of Business- to-Business Marketing. Journal of Business-to- Business Marketing , vol. 26, no. 1, p. 75–94, 2019. https://doi.or g/10.1080/1051712X.2019.1565142. [1 1 1] I. Zupic and T . Čater , “Bibliometric methods in management and or ganization,” Or ganizational r e- sear ch methods , vol. 18, no. 3, p. 429–472, 2015. https://doi.or g/10.1 177/10944281 14562629. Appendix Appendix A: Map of co-occurr ence network fr om abstract – Fig. 10 : Map of term co-occurrence network from ab- stract based on binary counting – Fig. 1 1 : Map of term co-occurrence network from ab- stract based on full counting 1 14 Informatica 47 (2023) 91–1 14 U.A. Bukar et al. Figure 10: Map of term co-occurrence network from abstract based on binary counting Figure 1 1: Map of term co-occurrence network from abstract based on full counting