Unleashing the power of internet of things and blockchain: A comprehensive analysis and future directions

As the fusion of the Internet of Things (IoT) and blockchain technology advances, it is increasingly shaping diverse fields. The potential of this convergence to fortify security, enhance privacy, and streamline operations has ignited considerable academic interest, resulting in an impressive body of literature. However, there is a noticeable scarcity of studies employing Latent Dirichlet Allocation (LDA) to dissect and categorize this field. This review paper endeavours to bridge this gap by meticulously analysing a dataset of 4455 journal articles drawn solely from the Scopus database, cantered around IoT and blockchain applications. Utilizing LDA, we have extracted 14 distinct topics from the collection, offering a broad view of the research themes in this interdisciplinary domain. Our exploration underscores an upswing in research pertaining to IoT and blockchain, emphasizing the rising prominence of this technological amalgamation. Among the most recurrent themes are IoT and blockchain integration in supply chain management and blockchain in healthcare data management and security, indicating the significant potential of this convergence to transform supply chains and secure healthcare data. Meanwhile, the less frequently discussed topics include access control and management in blockchain-based IoT systems and energy efficiency in wireless sensor networks using blockchain and IoT. To the best of our knowledge, this paper is the first to apply LDA in the context of IoT and blockchain research, providing unique perspectives on the existing literature. Moreover, our findings pave the way for proposed future research directions, stimulating further investigation into the less explored aspects and sustaining the growth of this dynamic field.


Introduction
The significant influence of technology has woven itself into the social fabric of society, ushering in transformative changes. Over the past few decades, technological advancements have instigated profound shifts in our lifestyle, reshaping the way we live, work, and communicate [1][2][3]. These developments have seeped into every aspect of our existence, creating an era characterized by heightened connectivity and data-centric decision-making J o u r n a l P r e -p r o o f [4]. The Internet of Things (IoT) stands as a testament to this technological revolution, representing a paradigm that morphs ordinary objects into smart, interconnected devices [5][6][7][8]. IoT encompasses an extensive network of physical objects, from simple household appliances such as thermostats and refrigerators, to complex industrial machinery. These items are equipped with sensors, software, and various technologies designed to collect and share data over the internet [4]. This digital connectivity engenders an ecosystem where the efficiency, convenience, and utility of these objects are remarkably improved. The ascendancy of IoT has been fueled by the digital revolution [9] and the pervasive reach of internet connectivity [10], leading to its exponential popularity. At its core, IoT embodies the concept of ubiquitous computing, which envisions computing as an integral, almost invisible part of our lives. It is designed to operate seamlessly in the background, assisting humans with a myriad of tasks and decisions [11]. IoT applications are not limited to a single sector but span a multitude of areas and demonstrate the flexibility and adaptability of this concept.
For example, in households, IoT is transforming mundane tasks through smart home systems [12].
According to [13], these systems manage energy consumption, automate home security, and even control home appliances, which can be monitored and controlled remotely. The introduction of IoT into smart homes has brought about a new level of convenience, control, and efficiency, and fundamentally altered the way people live. In the industrial sector, IoT is revolutionizing complex processes such as predictive maintenance in manufacturing.
Traditional maintenance strategies relied heavily on scheduled inspections and repairs, which were often inefficient and costly. With IoT, however, sensors embedded in machinery can monitor performance data in real-time and predict potential failures before they occur [14].
This not only reduces downtime but also extends the lifespan of the machinery, leading to significant cost savings and improved efficiency [5]. Furthermore, IoT is making considerable strides in sectors such as healthcare, transportation, and urban planning. [11] state that IoT devices like wearables and remote monitoring tools empower patients with more control over their health, providing real-time information. [15] argue that the technology has paved the way for intelligent transportation systems due its ability to enhance traffic management, reduce fuel consumption, and enable autonomous vehicles. In the context of urban planning, [16] highlight that IoT can serve as the foundation for the development of smart cities, where resources and services are optimized through data collected from citizens, devices, and assets.

J o u r n a l P r e -p r o o f
While IoT brings many benefits, it also introduces a range of challenges that need to be effectively addressed. The rapid development of the technology and the surge in interconnected devices have given rise to considerable privacy, security, and management issues. As such, IoT devices collect, process, and transmit vast volumes of data [5]. Often of a sensitive nature, this data can range from personal information like health records from smart wearables to critical business data from industrial machinery [11]. The sheer volume and variety of data these devices generate make data management significant challenges. It involves not just storing and processing this data but also ensuring its integrity, authenticity, and availability at all times [17]. Moreover, privacy concerns also arise as more and more personal information is collected and transmitted by IoT devices [18]. In a world where data has become a valuable commodity, the protection of personal information is paramount.
However, the use of IoT requires data from various aspects of a user's life to be constantly generated and transmitted, thereby potentially exposing users to privacy breaches [19].
Security is also one of the most pressing issues faced by IoT. The interconnected nature of IoT devices inherently makes them vulnerable to cyber-attacks [20]. Hackers can exploit weak security in one device to gain access to the network and compromise other connected devices [21]. Such breaches can have devastating consequences, especially in critical applications such as infrastructure or healthcare. Therefore, there is a need to overcome these challenges by introducing robust and innovative solutions that can safeguard the data, enhance privacy, and ensure the security of the IoT ecosystem [22]. This is where blockchain technology comes to the fore.
Blockchain is originally devised for the digital currency, Bitcoin, but it has evolved beyond its initial application, revolutionizing various sectors of the economy [22].
Blockchain is a type of distributed ledger technology that stores data across multiple systems in a way that is secure, transparent, and immutable [23]. The security feature of blockchain comes from its cryptographic algorithms and the decentralized nature of its network [24].
Each block in a blockchain is linked to then previous one through a cryptographic hash function, making it nearly impossible to alter data once it is recorded. Due to the fact that data is stored across a network of computers rather than a central server, it is difficult for hackers to compromise the system [25]. The transparent and immutable nature of blockchain makes it an ideal solution for data management and privacy concerns in IoT [26].
Transparency ensures that all transactions are open for verification by all participants, which enhances trust and collaboration [23]. Meanwhile, the immutability of blockchain ensures the J o u r n a l P r e -p r o o f integrity of data as it prevents any alteration of recorded data [27]. The combination of IoT and blockchain has potential applications in diverse areas, from creating secure, efficient supply chains to improving data privacy in smart homes.
The integration of IoT and blockchain has been attracting growing interest not only from industry practitioners looking to leverage these technologies for practical applications but also from the academic community. The burgeoning interest is underpinned by the recognition that the convergence of these two transformative technologies has the potential to catalyze innovation and technologies progress and shape the digital landscape of the future.
Academia has seen an increasing number of studies focusing on the intersection of IoT and blockchain. The surge in academic attention is indicative of the perceived significance and potential impact of these technologies when used in tandem. Numerous review studies have been conducted in this domain, each contributing to the cumulative understanding of the subject matter. For instance, [28] conduct an extensive examination of current blockchain protocols utilized in IoT networks and propose a classification of threat models addressed by blockchain technology. [29] review the possibilities of integrating blockchain and IoT to drive innovation in business models. Their findings reveal that incorporating blockchain into the IoT framework enables the establishment of a secure decentralized architecture, thereby enhancing existing businesses and facilitating the development of new business models by eliminating the reliance on third-trust parties.
Moreover, [30] investigate the current state of blockchain applications in the IoT domain and identify key research areas that enable blockchain to ensure security in large-scale distributed environments. The authors find that the potential of blockchain, specifically smart contracts, can improve the dependability and scalability of IoT applications by establishing trust for data and executed processes. [31] make a comprehensive review of how blockchain can be adapted to meet the unique demands of IoT, specifically in the development of blockchain-IoT applications, and find that blockchain can offer seamless authentication, data privacy, security, resistance to attacks, ease of deployment, and self-maintenance within the IoT context. Finally, [27] explore the role of blockchain in addressing data security concerns in IoT and highlight the challenges it faces in the IoT context. Despite this influx of research, a noticeable gap in the literature emerges upon closer inspection. While these review studies have provided valuable insights into the intersection of IoT and blockchain technology, none have utilized a comprehensive sample of journal articles and employed analytical methods, such as Latent Dirichlet Allocation (LDA). Theoretically, LDA is a sophisticated machine J o u r n a l P r e -p r o o f learning algorithm used for topic modeling, which allows for the discovery of abstract topics within a large collection of documents [32]. This method could offer a nuanced understanding of the main thematic structures of the body of literature concerning IoT and blockchain integration. Regrettably, the application of LDA in this context remains limited.
The current body of review studies, although valuable, might not fully capture the breadth and depth of this rapidly evolving field without the comprehensive and nuanced analysis that methods like LDA can provide [33]. Therefore, there is a compelling need for review studies employing such sophisticated methods to contribute to both theory and practice in this exciting intersection of IoT and blockchain technology.
The subsequent sections of this article are organized as follows: Section 2 presents the research method used in this study. Section 3 offers a summary of the review's findings. Section 4 analyzes the topics identified through the LDA approach. Finally, the article concludes by discussing the summary and limitations of the review.

Research method
This article employs a three-tiered hierarchical Bayesian model known as Latent Dirichlet Allocation (LDA) [34]. In theory, LDA comprises a set of algorithms designed to detect and tag the topics present in extensive text collections. It operates under the assumption that each document within the collection is a mixture of various topics, each characterized by a specific word distribution [35]. The objective of LDA is to uncover these latent topics by analyzing word patterns and co-occurrences across the documents. By utilizing LDA, analysts can generate K topics denoted as βK, representing probability distributions over terms in a set of texts using the vocabulary V. A notable strength of the LDA method lies in its ability to quickly and automatically extract relevant themes and patterns from vast volumes of text data. Consequently, we chose to use LDA in this study as it allows for a systematic and unbiased examination of an extensive corpus of literature, as demonstrated in previous research [33,36]. The tasks of statistical computing, graphical design, and natural language processing (NLP) were carried out utilizing the programming languages R and Python.

Literature selection
This study harnesses the power of text mining by extracting information exclusively from abstracts of chosen publications, specifically those addressing the synergy between IoT and blockchain technology. As in previous research utilizing LDA [37], our analysis primarily hinges on abstract-level scrutiny to yield insights into the intertwined nature of these two J o u r n a l P r e -p r o o f groundbreaking technologies. This method ensures that we encapsulate relevant information concerning the overlapping applications of IoT and blockchain technology. By dissecting abstracts, we are able to identify a broad spectrum of topics, trends, and insights that might otherwise be concealed. As a result, our examination offers a more comprehensive exploration of the potential advantages and challenges intertwined with the intersection of IoT and blockchain technology. To gather pertinent publications, we turned to the wellregarded Scopus database, which is renowned for its inclusive coverage of academic studies.
On April 1st, 2023, we conducted searches in Scopus using the following search query: ( "internet of things" OR iot OR rfid OR wsn OR "wireless sensor network*" OR gps OR actuator* OR sensor* ) AND ( blockchain* OR "block-chain*" OR "block chain*" ) [4].
The terms within the first set of parentheses are used to find publications related to IoT.
These terms encompass the broad category of IoT, including specific aspects such as Radio-Frequency Identification (RFID), Wireless Sensor Network (WSN), GPS, actuators, and sensors. The AND operator is used to ensure that the search results include both the IoTrelated and the blockchain-related terms. The second set of parentheses are used to find publications related to blockchain technology. The Asterisk (*) is a wildcard symbol that allows the search to include variations of the term. For example, "blockchain*" would include "blockchain", "blockchains", "blockchain-based", and so on. By using these two sets of terms together with the AND operator, the search query will return only those publications that contain at least one term from each set, thereby ensuring that the publications are relevant to the interplay between IoT and blockchain technology. In line with previous research [38], we restricted our investigation to articles written in English to ensure the academic integrity of the information procured. We aimed for an exhaustive exploration of To enhance the precision our analysis, a tailored set of stop words was formulated by supplementing the default Gensim list with additional irrelevant terms pertaining to our specific context. Finally, the Gensim library was employed to disassemble phrases into discrete words and assign them distinctive identifiers (IDs). By adhering to this systematic approach, we could not only enumerate the frequency of individual words across the texts but also assess their relative significance within the context of blockchain and IoT interplay.

Development of the LDA model and identification of optimal K number of topics
The construction of an effective lexicon is a pivotal step in developing the LDA model and deciphering themes from the texts concerning IoT and blockchain. The 'id2word' function from the Gensim package allows for the creation of a vectorized bag of words swiftly and effortlessly. Subsequently, the Mallet toolset, known for its versatility in clustering, topic modeling, and document classification, was employed to build the LDA model [39]. Mallet is a valuable instrument for analyzing data and gleaning insights. The modeling process necessitates various configurations. For this study, Mallet was used to execute multiple simulated LDA models, each comprising different numbers of topics. The optimal number of topics was determined using the topic coherence score, which evaluates the coherence of a topic by scrutinizing the relationships among its constituent words.
In essence, this metric assigns a numerical score to each topic that echoes the extent to which its words correlate and form a cohesive theme. A higher coherence score signifies a stronger interrelation among the words within a topic and indicate their significant contribution to the theme. This is desirable for accurate and precise topic modeling. Figure 1 illustrates the coherence scores produced by the unsupervised learning system. Guided by the principle of coherence score, the most effective LDA model would be one that possesses the highest and most consistent score because this indicates greater semantic coherence within each topic. The outcomes reveal that after implementing 14 topics, the model reaches a mean J o u r n a l P r e -p r o o f value of 0.4393, suggesting that incorporating more than 14 topics would not yield additional valuable insights. Consequently, the model with 14 topics was selected as the ideal choice for analysis based on the coherence scores generated during the LDA modeling process. Table 1 presents the coherence scores for all the assessed topic numbers.

Identification of topics
The LDA model is a type of probabilistic model utilized to unearth topics from a collection of documents [40], in our case, those related to the integration of IoT and blockchain technology. In the graphical representation shown in Figure 2, rectangles are depicted as replicates, with 'M' representing documents and 'N' indicating the occurrence of a topic within each document. The distribution of observed words, marked as 'w', is predicated upon the topic distribution, signified as 'z'. In this model, 'β' denotes the distribution of words across topics, 'θ' signifies the distribution of topics across documents, and 'α' illustrates the distribution of words within individual topics. We employed the LDA model to identify the frequency at which the fourteen topics gleaned from the literature were discussed in the chosen journal articles. We adopted the semantic coherence approach to quantify the frequency of topic-associated terms present in the abstract of each article. Through an inductive process based on the semantic coherence score, two researchers independently curated a set of articles for each topic. This allowed the LDA model to highlight latent topics unique to each document, along with their relevance and frequency across the texts, thereby shedding light on the current state of research regarding the intersection of IoT and blockchain technology. The LDA model was derived from all abstracts and analyzed using several Python packages. PyLDAvis was employed to ascertain the average distance between topics and the ten most significant terms within our dataset. Additionally, the Matplotlib library was used to graphically depict the research findings and enhance the clarity and understanding of the results.

Bibliometric analysis
To gain deeper insights into the significance of the chosen journal articles on IoT and blockchain technology, a bibliometric analysis was carried out. We employed the bibliometric R package to execute the methodology proposed by [41]. This tool simplifies the process of discovering links between academic papers, thereby enabling researchers to thoroughly understand the underlying networks and themes within the data. Our study J o u r n a l P r e -p r o o f primarily aimed to conduct performance analysis and scientific mapping, both of which were achieved through the use of bibliometric methods. As elaborated by [42], the former allows for intricate scrutiny of scholarly collaborations and research output, while the latter facilitates the comprehension of the genesis and development of a specific research domainin this case, the intersection of IoT and blockchain technology.

Descriptive results
The objective of the bibliometric analysis was to delve into the prominent academic journals exploring the interplay between IoT and blockchain technology. synergizing in small to medium-sized groups to address the multifaceted aspects of IoT and blockchain integration. It may also suggest that the field is interdisciplinary in nature, with scholars from diverse backgrounds and areas of expertise uniting to tackle complex challenges. In analyzing the robust and dynamic landscape of research on the convergence of IoT   particularly in situations where traditional data management techniques prove to be inadequate or unsafe [43][44][45][46]. The considerable attention dedicated to this topic underscores J o u r n a l P r e -p r o o f the significance of developing innovative solutions that leverage blockchain technology to enhance the privacy, reliability, and security of healthcare data management [43].
Conversely, the central role of Topic 3 in the literature reflects the increasing interest in developing more effective and seamless supply chain management solutions using IoT and blockchain technology. As global trade continues to grow and the demand for efficient logistics increases, supply chain management becomes a critical challenge. Consequently, the integration of IoT and blockchain in this area offers the potential to optimize supply chain processes by improving traceability, reducing operational costs, and enhancing overall supply chain security [22]. The growing interest in this topic emphasizes the importance of exploring innovative IoT and blockchain solutions to elevate the efficiency, reliability, and transparency of supply chain management [5,47].
Utilizing PyLDAvis, a Python package developed by [48], the significance of the  providing a secure and decentralized mechanism to manage access and permissions across multiple devices [23]. This decentralization aspect of blockchain also plays a crucial role in Topic 14, where the use of IoT devices in wireless sensor networks could benefit from blockchain technology for improving energy efficiency [49]. By enabling peer-to-peer energy transactions, blockchain could potentially minimize losses associated with centralized energy distribution, thereby enhancing overall energy efficiency [50].

Security and authentication in blockchain-based IoT networks
Topic 1 is primarily centered around the concepts of blockchain technology and its applications within the realm of IoT security. The key terms associated with this topic, such as scheme, security, authentication, device, blockchain, IoT, attack, key, protocol, and data, attacks. Finally, [55] propose an effective and innovative security scheme for IoT application by combining the advantages of blockchain, random number generation, and dynamic key generation. In summary, Topic 1 sheds light on the use of blockchain technology as a key enabler for security in IoT. The topic emphasizes the need for robust, efficient, and scalable security mechanisms in various IoT domains, from IoD and IIoT to Wireless Body Area Networks (WBANs), and presents blockchain as a potential solution to these changes. As such, this topic appears to be an important area of focus for researchers aiming to enhance the security and trustworthiness of IoT networks. Based on the discussion of Topic 1, emerging future research directions might include: 1. Exploring more efficient blockchain consensus algorithms specifically designed for IoT environments to improve system performance and security.

Blockchain and IoT integration in supply chain management
The third topic is labeled "Blockchain and IoT integration in supply chain management", and it focuses on the fusion of blockchain technology, IoT, and supply chain management.
Technology, blockchain, chain, supply, digital, industry, AI, IoT, are a few of the most significant terms related to this topic. More specifically, these terms denote a distinct

Federated learning and blockchain in smart systems
Topic 4 is titled "Federated learning and blockchain in smart systems", and it predominately focuses on the integration of federated learning, blockchain technology, and intelligent systems. Conceptually, federated learning represents a machine learning approach that allows for decentralized learning, where the model learns from data located on different devices or servers without the need to centrally collect this data. The integration of blockchain technology in such systems could provide additional security and privacy benefits, making this combination especially relevant for applications dealing with sensitive or private data.

Blockchain in IoT and IIoT for data security and privacy
Topic 5 is labeled "Blockchain in IoT and IIoT for data security and privacy," and it concentrates on the utilization of blockchain technology within the IoT and the Industrial Internet of Things (IIoT). This includes a focus on ensuring data security, privacy, and managing trust relationships. Key terms such as data, smart, system, contract, service, user, J o u r n a l P r e -p r o o f trust, device, management, privacy, security, IIoT, sharing, application, network, mechanism, platform, solution, technology, decentralized, and framework all align with the topic. From the papers reviewed, it is clear that there is a significant interest in using blockchain technology to improve trust management in IoT networks [74], to bolster security systems in smart homes [75], and to create a real-time interactive platform based on publish-subscribe mechanism [76]. Another focus of research on this topic is data trading and sharing. For instance, blockchain has been used to facilitate decentralized data trading [77], to manage the cost of IoT sensor data storage [78], and to incentivize vehicular crowdsensing activities [79].
The use of smart contracts is another recurring theme. This includes using smart contracts for data commodity transactions in the IIoT [80], fuzzing smart contracts for TOD vulnerability detection [81], and proposing a blockchain-based privacy-preserving reputation framework for participatory sensing systems [82]. Several studies also concentrate on data privacy, with some proposing a decentralized personal data store based on Ethereum to achieve GDPR compliance [83] and a blockchain-enabled antileakage sharing protection scheme for undisclosed IIoT vulnerabilities [84]. Others focus on operational data security in industrial control systems [85] and a fair, secure, and trusted decentralized IIoT data marketplace enabled by blockchain [86]. Finally, scholars explore the intersection of semantic knowledge management, blockchain, and privacy in IoT applications [87]. In short, Topic 5 sheds light on the potential of blockchain technology in enhancing security, privacy, and trust management in IoT and IIoT systems. As such, future research directions might include: 1. Developing predictive models that can anticipate and mitigate security vulnerabilities in blockchain-IoT systems.
2. Investigating the role of quantum computing in enhancing blockchain security and data privacy in IoT and IIoT.
3. Evaluating the potential of federated blockchain models for improved data sharing and privacy in IoT and IIoT environments. 4. Applying AI and machine learning for the automation of smart contracts to increase efficiency and security in IoT and IIoT data exchanges. 5. Exploring how blockchain can facilitate edge computing in IoT and IIoT for better data security, privacy, and decentralized processing. Developing adaptive blockchain algorithms that can dynamically adjust to the evolving security requirements of IoT and IIoT applications. 8. Assessing the application of blockchain in ensuring data integrity and traceability in critical IoT and IIoT systems like healthcare or industrial control systems. 9. Designing decentralized identity solutions using blockchain in IoT and IIoT to enhance user privacy and control over personal data.

Blockchain in healthcare data management and security
Topic 8 deals with the application of blockchain technology in the healthcare industry for secure data management. Keywords such as data, blockchain, system, IoT, security, healthcare, access, and privacy indicate that the core theme of the topic centers on ensuring the security, privacy, and controlled access of healthcare data within IoT systems by leveraging blockchain technology. One of the key areas under this topic is the control and management of personal health data. For example, [101] introduce an IoT-based configurable blockchain for mHealth data, offering privacy protection, user control, and HIPAA compliance, thereby enabling personalized healthcare systems with secure data storage and analysis. [102] introduce a scheme that utilizes blockchain technology to protect the privacy of medical data during sharing. By employing K-anonymity, searchable encryption, and Hyperledger Fabric, the scheme ensures secure access control and confidentiality while also demonstrating practical scalability and performance. Several studies also discuss the secure storage and sharing of healthcare data. For example, [103] present a novel framework that leverages WBAN and blockchain technology to guarantee the confidentiality, security, and authenticity of health data. By integrating sensor devices, cloud storage, and blockchain, the proposed framework enables secure data transmission and storage, effectively addressing privacy and security concerns in healthcare. The system aims to benefit patients, healthcare providers, and health insurance providers by enabling informed self-care, remote patient J o u r n a l P r e -p r o o f monitoring, and preventing fraudulent claims. Similarly, [102] propose a model for a secure and decentralized medical information system using blockchain technology, enabling safe storage and sharing of medical data, real-time data collection during surgery, anonymous data sharing, and implementation with Hyperledger Fabric. Moreover, the studies also explore the integration of blockchain with other technologies for enhancing healthcare data security. For example, [104] introduce a blockchain-based access control system using smart contracts, providing a secure and trustworthy method for sharing electronic health records, addressing the challenges of third-party dependence and ensuring privacy in healthcare data sharing. 6. Identifying and addressing potential cybersecurity threats for advanced network systems using blockchain technology.

Blockchain in supply chain and transportation systems
Topic 11 concentrates on the deployment and potential advantages of blockchain technology within the realms of supply chains, food systems, and transportation logistics. The terms such as vehicle, system, chain, blockchain, supply, food, data, and IoT affirm the exploration and evaluation of blockchain applications in these particular domains. An important aspect within this topic is the application of blockchain technology in supply chain management. Studies like [114], [115] and [116] offer insights into how blockchain can promote transparency, balance, and agility in supply chains, contributing to their overall efficiency and robustness.
Another significant theme emerging from this cluster is the application of blockchain in food systems, a notion underlined by representative terms such as food, traceability, quality, and safety. This theme is represented in studies like [117] and [118], which propose blockchain-

Blockchain and IoT applications in smart cities
Topic 12 covers the integration and potential advantages of blockchain and IoT technologies within the realm of smart city solutions. The recurring terms, including blockchain, IoT, smart, security, application, device, system, technology, network, city, challenge, transaction, privacy, and solution, reveal the focus of the research on enhancing security, privacy, and overall operational efficiency of IoT-based smart city applications using blockchain technology. Key areas of research within this topic include the examination of security and privacy concerns related to IoT systems in the context of smart cities. [121] examine the security concerns in smart cities and the implications of quantum computers on blockchainbased applications, presenting a blockchain framework tailored for smart cities and exploring potential solutions to enhance security in the face of quantum threats. In addition, [122] propose a novel approach using a blockchain-defined network and a grey wolf-optimized modular neural network to enhance security in smart environments, effectively addressing security, privacy, and confidentiality issues. The results show that the proposed system J o u r n a l P r e -p r o o f achieves exceptional security (99.12%), improved efficiency, and low latency compared to other neural networks, such as multi-layer perceptron and deep learning networks.
Studies like [123] and [124] discuss how IoT and blockchain can support secure, efficient data communication within smart cities. Moreover, some papers, such as [125], present the Blockchain-of-Blockchains (BoBs), a hierarchical blockchain-based platform designed to address data management challenges, ensure data integrity, and enable interoperability in

Evaluating real-world applications and challenges of implementing blockchain in
WSNs.

Conclusions
Digitalization has become an integral part of modern human life, impacting every aspect of our daily activities. Although the intersection of blockchain and IoT is burgeoning with potential, there seems to be a dearth of data-driven reviews exploring this synergy, specifically using sophisticated methods like Latent Dirichlet Allocation (LDA). As a powerful form of topic modeling, LDA can provide valuable insights by identifying latent patterns and topics within vast amounts of text data. As a data-driven tool, it has the potential to explore the nuances and themes in the emerging blockchain and IoT literature, offering a comprehensive overview that could inform researchers, technologists, and decision-makers. The lesser representation of Topics 2, 14, and 9 in the corpus might initially seem surprising, but it also signals potential opportunities for deeper exploration and greater innovation in these areas. Topic 2, which revolves around "Access control and management in blockchain-based IoT systems," has received relatively less attention. A possible explanation for this could be the inherently complex nature of devising efficient and secure access control mechanisms in decentralized IoT networks. It might also point to a greater J o u r n a l P r e -p r o o f focus on more immediate, overarching security issues associated with IoT and blockchain.
However, the underemphasis on this topic could overlook the nuanced and crucial role of access control in the broader security architecture, which can have profound implications for user privacy and data integrity. In addition, Topic 14, "Blockchain, IoT, and energy efficiency in wireless sensor networks (WSN)," might not have received as much focus due to the highly specialized nature of the topic. Integrating energy efficiency with blockchain and WSN in IoT might require interdisciplinary expertise that bridges not only information technology and blockchain, but also energy systems and networking. Moreover, the topic's practical applications might be seen as confined to specific industries or sectors, leading to a lower general interest. Therefore, the unique combination of these technologies could offer unprecedented opportunities to improve the sustainability and efficiency of IoT systems, an aspect of growing importance in the age of environmental consciousness. Finally, the lower emphasis on Topic 9, "Blockchain in energy systems and trading," could suggest that the integration of blockchain into the energy sector is still in its nascent stage. This might be due to regulatory hurdles, technical challenges, and the historical inertia of traditional energy systems. With global climate commitments and the increasing feasibility of decentralized energy systems, there may be substantial opportunities in the near future for blockchain to revolutionize this sector, thereby warranting more scholarly attention.
Although this study is comprehensive, it is not without its limitations. Firstly, the reliance on the LDA method as a powerful tool for topic modeling may not perfectly distinguish between topics, especially when they are closely related or overlapping.
Therefore, this may lead to a lack of precision in topic identification. Secondly, our review was confined to English language research articles indexed by the Scopus database.
Consequently, the review may exclude relevant research published in other languages or in non-indexed journals, potentially introducing a language and indexing bias. Moreover, the time frame of the study extends only up until early 2023, meaning that the latest developments and trends may not be captured. Finally, given the fast-paced evolution of blockchain and IoT technologies, certain emerging topics and novel applications might not have been fully incorporated in the current review. As a result, there is a need for continuous updating and reviewing. Despite these limitations, we believe our study provides valuable insights and a comprehensive overview of the research landscape at the convergence of blockchain and IoT.