GSTLB: A Secure GPS-Enabled Software-Defined Network Leveraging Transfer Learning and Blockchain for Industrial IoT (IIoT)

GSTLB: A Secure GPS-Enabled Software-Defined Network Leveraging Transfer Learning and Blockchain for Industrial IoT (IIoT) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article GSTLB: A Secure GPS-Enabled Software-Defined Network Leveraging Transfer Learning and Blockchain for Industrial IoT (IIoT) BAZEER AHAMED BAGRUDEEN This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5963243/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract IIoT sensors bridge the gap between industry and technology by assisting in intelligent automation of the sector. The proposed method uses GPS to track assets and identify products for customers. It also uses TL (Transfer Learning) to identify machine problems, which alerts for immediate service. Using the 5G network greatly facilitates faster communication and surveillance control. All of the data from these various applications, including GPS location tracking, surveillance, and alerting, are stored in a block chain where malicious data can be added with the original data or eves dropping is possible, which could interfere with the industry's normal operations. Plot-based features and CNN-based feature extraction are both used to get around this, and the network's efficiency and security are increased by this hybrid approach of separating dangerous input from the original data. When using the block chain methodology, data is compressed to make it unified because Network Slicing is the primary concept utilized in industry and SDN (Software Defined Network) is used as a centralized one for communication. According to the simulation results, the suggested methodology improves IIoT network security, speeds up data transmission, and conducts ongoing product and machine inspections, all of which significantly aid in expanding the supply chain. SDN Block Chain Technology Transfer Learning IIoT network security Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Smart automation's incorporation of the Industrial Internet of Things (IIoT) closes the gap between cutting-edge technology and conventional industry methods. In order to improve communication, surveillance, and operational efficiency, this project uses 5G networks, transfer learning (TL), and blockchain technology to present a novel approach to asset tracking and machine problem identification using GPS and sophisticated analytics. Customers can keep an eye on products all the way through the supply chain with real-time GPS monitoring, and sensors can identify machine problems and send out maintenance notifications to save downtime. 5G's fast speeds facilitate data analysis and real-time monitoring. Blockchain technology guards against malicious data insertion and eavesdropping, protecting data from GPS monitoring, surveillance, and alarms. Plot-based feature extraction and Convolutional Neural Network (CNN)-based feature extraction are combined in a hybrid security approach to better separate harmful from valid data, hence improving security. Software-Defined Networking (SDN) offers centralized network control, while network slicing maximizes resource allocation. The blockchain compresses data to make it easier to store and retrieve. According to simulation results, this approach greatly improves IIoT network security, permits faster data transfer, and makes it easier to conduct ongoing machine and product inspections, all of which increase supply chain efficiency. A more effective and safe industrial ecosystem is made possible by this all-encompassing approach, which modernizes industrial operations and guarantees improved asset tracking, machine maintenance, and data protection. Literature review IIoT focuses on monitoring the PH level [ 1 ] and temperature of the contaminated water and when it encounters the water flows with the highest contamination level, the water flow valves were controlled and that water was sent to the water treatment plant and processed and sent to the plant and these data such as monitoring of PH evel and temperature of water by sensor, excess flow of waste water everything is monitored. The data is collected utilising blockchain [ 2 ], where the blockchain technology aids in the expansion of the supply chain.H1: Ownership concentrationhas no effect on firm-level stock return. They suggested a heuristic algorithm using AI and the idea of network slicing to replace the old way of utilising the physical network, and they created the concept of network slicing orchestration system[ 3 ], so that various applications may utilise their specialised network. In IIOT, all of these processes are active at the same time. When considering the of IOT Botnet in IIOT[4,] for trying to overcome an approach called graph based feature which can help in differentiating or differentiating both malicious and benign samples, for gathering data samples among the 3 techniques available in botnet detection, dynamic analysis methods and produces an accuracy of 98.1% and 91.99% in botnet detection Surveillance [ 5 ] has grown increasingly important in recent years, yet cameras can still identify people and objects. To boost trust in IIOT-based monitoring, AI has been applied, which aids in the detection of dangerous objects, as well as the usage of a violence detection network. After the harmful object is recognised, an alert is sent to the appropriate departments, and CNN is used for frame identification. Because of numerous entities such as packaging units, hubs, and manufacturing units, IIOT substantially lags in data integration, therefore data is merged via block chain technology [ 6 ], where data transparency is assured safely and hostile device attacks are avoided The technological advances in IIoT [ 7 ] lead to the development of light weight devices with less energy consumption and longer range of wireless connection, because of this there has been a development such as smart airport, where enormous sensors has incorporated and enabled various wireless technologies such as Wi-Fi, and so on, which enable data collection and adaptation of system to the varying environment. The use of Industrial Internet of Things (IIoT) in smart airports has brought about numerous benefits such as improved communication, business processes, and efficiency. However, it has also introduced new vulnerabilities that can be exploited by cyber attackers to compromise both the digital infrastructure and physical assets of the airport. To address this issue, a new security-oriented IIoT testbed called SAir-IIoT has been developed. This testbed comprises multiple IIoT devices and communication protocols that are organized into distinct zones and can be remotely accessed as a service. The purpose of this paper is to present SAir-IIoT as a novel solution to the security challenges posed by IIoT in smart airports. Our comparison of SAir-IIoT with other IIoT-based testbeds reveals its complexity and effectiveness in evaluating new cyber security methods. SAir-IIoT is compared with other IIoT-based testbeds in order to demonstrate its complexity and effectiveness when evaluating new cyber security methods. In conclusion, we compare SAir-IIoT with other IIoT-based testbeds, revealing both its complexity and effectiveness. As a final result, we compare SAir-IIoT with other IIoT-based testbeds, demonstrating the complexity and effectiveness of this testbed with respect to evaluating new cyber security approaches. Our final analysis compares SAir-IIoT with other IIoT-based testbeds in order to reveal its complexity and suitability for the evaluation of new cyber security methods. Our final evaluation of SAir-IIoT reveals its complexity and effectiveness in evaluating new cyber security techniques. In addition, we compare SAir-IIoT to other IIoT-based testbeds to demonstrate its versatility and effectiveness in assessing new cybersecurity techniques. We finalize our study by comparing SAir-IIoT with other IIoT-based testbeds, in order to demonstrate the complexity and effectiveness of the testbed when it comes to evaluating new cyber security methods. In addition, we have compared SAir-IIoT with other IIoT-based testbeds, indicating its complexity and efficacy as a tool for testing new cyber security methodologies. Finally, we assess SAir-IIoT compared with other IIoT-based testbeds, revealing its complexity and effectiveness for the evaluation of new cyber security technologies. It is difficult to address the combined optimization problem of a 5G-inspired IIoT-MEC interactive network, which aims to maximize MNO income and minimize IIoT operators' economic cost. In order to increase MNO income while maintaining reasonable service charges for IIoT mobile devices (MDs), the author [ 8 ] of this letter suggests a dynamic pricing model for IIoT-MEC networks. A discrete finite Markov decision process (MDP) is used as the initial model for the dynamic pricing issue. After that, this problem is resolved using the Q-learning technique. The findings demonstrate that the suggested dynamic pricing method may greatly increase MNOs' income and lower the economic cost of IIoT operators. The paper [ 9 ] that proposes a novel method for preserving sensitive data in IIoT operations. The Industrial Internet of Things (IIoT) is being increasingly used in manufacturing and related industries, and Machine Learning (ML) techniques are commonly used to analyze the collected data. However, ML applications require decrypted data to perform efficiently, which raises concerns about data privacy and security. To address these concerns, the paper proposes a hybrid method of Generative Adversarial Networks (GAN) and Differential Privacy (DP) to preserve sensitive data in IIoT operations. The proposed method aims to sustain IIoT data privacy with minimal accuracy loss and additional computational costs. The paper uses publicly available datasets and a realistic IIoT dataset collected from a confectionery production process to demonstrate the efficiency of the proposed approach. The paper employs privacy assessment metrics and shows that the proposed method preserves the privacy of the data while maintaining the accuracy of the algorithms used for analysis. Overall, the method aims to prevent the production of hidden sensitive data from the sub-feature sets and ensure privacy protection for hidden sensitive data in IIoT operations. Modern urbanization [ 10 ] and smart cities have undergone a transformation thanks to recent advances in the Industrial Internet of Things (IIoT). Even though IIoT data contains a wealth of interesting events and objects, analyzing a sizable volume of IIoT data and creating predictions in real time are difficult. Recent developments in AI enable analysis of this enormous volume of IIoT data and the generation of insights for subsequent decision-making processes. In this post, we outline some essential features of IIoT data with AI support for monitoring smart cities. In order to gather events and objects from IIoT data in real time, we first integrated an AI-enabled IIoT framework with a crowdsourcing application that supports human intelligence. In order to automatically classify the collected events and objects and provide analytics, reports, and warnings from the IIoT data in real time, we have merged numerous AI algorithms that can operate on distributed edge and cloud nodes. Two scenarios can be used the results in. In the first instance, the smart city authority can validate the events that were processed by AI and allocate them to the proper authority for event management. In the second scenario, AI algorithms are permitted to communicate with people or the IIoT to complete tasks. The implementation specifics for the aforementioned situations as well as the test findings will be presented in the final section. The framework has the potential to be used in a smart city, according to the test findings. Due to the dynamic and open nature of the Industrial Internet of Things (IIoT) [ 11 ] ecosystem, pictures produced by smart cameras and sensors are significantly at risk when sent over a public network. A practical strategy for protecting IIoT digital photos is encryption. This article discusses a framework for picture security that makes use of DNA cryptography and chaotic maps. The suggested approach generates three keys using a multilayer combination of a tent, circle, Chebyshev, and 3-D logistic map. These keys are utilized to build a key image on which DNA XOR operation is carried out to obtain the encrypted image, decide the rule to do DNA encoding-decoding on the subblocks, and perform row-column rotation of the subblocks. The result analysis of the suggested scheme shows that its average NPCR (99.6566%), UACI (33.4588%), Entropy (7.9971), and bigger key-space of 10195 are superior to those of the existing schemes and more resilient to various assaults. Increased security, transparency, and traceability are promised as a result of integrating [ 12 ] blockchain-IIoT into industrial operations. This development, however, runs into serious storage and scalability problems with current blockchain systems. A complete copy of the ledger is kept by each peer in the blockchain network, and it is updated by consensus. This complete replication strategy puts a pressure on the peers' storage space and would rapidly exceed the storage capacity of IIoT devices with limited resources. In the literature, a number of methods utilizing compression, summarization, or various storing systems have been suggested. In the last several years, there has been a lot of research done on the utilization of cloud resources for blockchain storage. Block selection is still a significant issue with cloud resources and blockchain integration, though. In order to solve the block selection problem, which entails choosing the blocks to be moved to the cloud, this research suggests a deep reinforcement learning (DRL) method. By transforming the multi-objective optimization of block selection into a Markov decision process (MDP), we suggest a DRL technique to tackle our problem. We create a mock blockchain environment to practise and test our suggested DRL methodology. The block selection issue is solved using two DRL algorithms, Advantage Actor-Critic (A2C) and Proximal Policy Optimization (PPO), and their performance benefits are examined. In comparison to the complete replication strategy used by traditional blockchain systems, PPO and A2C achieve storage reductions on the blockchain peer of 47.8% and 42.9%, respectively. The run-time difference between the slowest DRL algorithm, A2C, and the benchmark evolutionary algorithms employed in prior publications is only 7.2 times, validating the benefits of the DRL algorithms. The outcomes of the simulation further demonstrate that our DRL algorithms offer a flexible and dynamic response to the time-critical blockchain-IIoT context. The industrial Internet of Things (IIoT) [ 13 ] network places a high value on data, therefore trust and data security are two of the main issues. In order to safeguard data transmitted via wireless communication, contend with unauthorized entities, and guarantee data integrity, we create a cloud-integrated 5G-IIoT network architecture supported by a three-party authenticated key exchange (AKE) protocol with privacy-preserving. In addition, we create a trust model based on the Dempster-Shafer theory to evaluate the reliability of information gathered by smart devices/sensor nodes. Our scheme has withstood several well-known assaults, according to security assessments, in the IIoT environment. Using a tool for automated validation of internet security protocols and apps, we also examined the correctness of our plan. Additionally, the experimental and performance assessment findings demonstrate that the suggested strategy outperforms other works in terms of accuracy, latency, trust, and throughput. This article discusses the challenges associated with handling delay-sensitive [ 14 ] and compute-intensive workloads in the context of Industrial Internet of Things (IIoT) systems, and proposes a solution based on mobile edge computing (MEC) to offload these tasks for processing. However, offloading tasks can lead to higher energy consumption, which can impact battery life, and the communication delay between IIoT devices and MEC can also impact performance. To address these issues, the article proposes a novel delay-aware energy-efficient (DAEE) online offloading algorithm that can adaptively offload more tasks when the network quality is good, while delaying transmission when connectivity is poor but ensuring that task deadlines are not violated. The article also discusses the theoretical underpinnings of the proposed algorithm and provides simulation results to demonstrate its effectiveness in minimizing energy consumption and maintaining low latency, especially for delay-sensitive and compute-intensive tasks. This article [ 15 ] discusses the security and privacy issues that arise when devices from different management domains in the Industrial Internet of Things (IIoT) communicate with each other. Existing authentication schemes have limitations such as a single point of failure in a trusted center, high certificate management costs, and low authentication efficiency. To address these issues, the article proposes an efficient and anonymous cross-domain authentication scheme based on blockchain technology. The scheme uses dynamic accumulator technology and combines it with blockchain to achieve fast authentication while ensuring device anonymity to prevent link ability of identities. The article also discusses the security analysis and performance evaluation of the proposed scheme, which demonstrates its ability to resist common attacks and its feasibility and efficiency in practice. Overall, the article proposes a novel approach to secure cross-domain communication in the IIoT using blockchain technology. This paper [ 16 ] discusses the challenges of extracting value from the vast amounts of data produced by Industrial Internet of Things (IIoT) infrastructure, which is increasingly integrated with artificial intelligence (AI) and machine learning (ML) solutions. Although digital marketplaces have emerged to enable data owners to monetize their data, concerns such as privacy, security, and fair payment settlement have hindered their adoption. Additionally, the current centralized platforms are controlled by large multinational corporations and lack transparency between buyers and sellers. To address these challenges, the article proposes a decentralized platform for a digital data marketplace for IoT data that leverages a decentralized data streaming network for reliable and fault-tolerant hosting of IoT data. The platform also ensures fair trading, data storage, and delivery in a privacy-preserving manner, and calculates trust metrics for actors in the network. The feasibility of the proposed platform is studied through the development of an open-source library using Hyperledger Fabric and VerneMQ, which is tested on a real-time Google cloud platform for throughput, overheads, and scalability. Overall, the article proposes a novel approach to enable fair and transparent monetization of IoT data in a decentralized manner. This paper discusses [ 17 ] the importance of physics-informed learning for safety in Industrial Internet of Things (IIoT) systems and proposes a new method for IIoT intrusion detection called MTID. The conventional intrusion detection methods require auxiliary equipment and are not versatile enough for general IIoT systems. MTID is based on the temperature fingerprint of the microcontroller unit (MCU) chip and uses a self-encoder-based intrusion detection model to identify the security status of the nodes. The temperature residuals dataset is constructed by analyzing the relationship between the temperature sequence and the computational complexity of the node. The proposed method is designed to be applicable under the diversified deployment environment of IIoT systems, and an online incremental training method is developed and applied to ensure the model's applicability. Experimental analysis using Raspberry Pi 4B shows that MTID achieves an accuracy of 89% for intrusion detection, which demonstrates the feasibility of the intrusion detection method based on MCU temperature. Proposed Methodology A number of cutting-edge technologies are integrated into the suggested methodology for the "GPS Enabled Secured Software Defined Network using Transfer Learning and Blockchain Technology for Industrial IoT (IIoT)" in order to accomplish safe, effective, and dependable industrial operations. GPS technology makes it possible to track machines and assets in real time, offering location-based services and ongoing monitoring. By applying transfer learning, pre-trained models are modified to fit the unique IIoT context, improving the capacity to recognize machine problems and efficiently initiate maintenance alerts. By safely storing GPS tracking, surveillance, and alert data and guarding against eavesdropping and malicious data insertion, blockchain technology guarantees data integrity. A hybrid security approach uses both plot-based and CNN-based feature extraction to differentiate between malicious and legitimate data. Network slicing optimizes resource allocation by establishing virtual networks customized to particular applications and requirements, while Software Defined Networking (SDN) offers centralized network control, facilitating effective administration and integration of various IIoT data streams. Four types of services are provided by the suggested method: The supplier can ensure payment, storage, quality control, and ordering. The worker gains knowledge in planning, reporting, human resources, and goal attainment. Order planning, logistics, and services all benefit customers. Additionally, machines can be programmed to understand uptime and downtime. Blockchain technology compresses data for effective storage and retrieval, and unified product data management improves customer choices and data management in general. DB-CGAN algorithms are used in conjunction with CNNs to improve classification performance, and CNN-based feature extraction is used for real-time data monitoring to identify and eliminate fraudulent inputs, guaranteeing data integrity. According to simulation results, the suggested methodology greatly improves supply chain efficiency by enabling faster data transfer, enhancing IIoT network security, and facilitating ongoing product and machine inspection. Using algorithms like SMOTE, ADASYN, MENGNETO, and DB-CGAN, several classifiers (RF, SVM, DNN, and CNN) are compared; CNN-based DB-CGAN yields the best results. Additionally, various CNN architectures (e.g., GoogleNet, ResNet-50) are assessed for accuracy and training time; GoogleNet is found to be the most effective. The suggested solution seeks to improve the overall security, effectiveness, and dependability of IIoT applications by integrating these techniques, guaranteeing reliable asset tracking, effective machine maintenance, and safe data handling. Continuous machine monitoring, as shown in Fig. 1 equipment or products, is the second type of service. The TL[ 7 ] uses its experience to determine when the downtime and uptime will occur; if there is a deviation, the maintenance manager will be informed and the necessary action will be taken. Additionally, there is a chance that malicious data may be added unintentionally or intentionally when using IIoT. For this reason, data will be continuously monitored by a Convolutional Neural Network and a graph feature-based method. If any deviations are discovered, the data is eliminated and the network is restored to normal., This is the third type of service and the fourth is SDN [ 8 ] service; however, because of the numerous components in the industry, including hubs, production, and packaging, data integration becomes difficult because of network slicing. The block chain is used to combine all of these data, and the SDN service keeps an eye on them for optimal security. The supply chain's growth is facilitated by this data integration. Through sophisticated data management and real-time monitoring, the incorporation of Industrial Internet of Things (IIoT) technologies into industrial processes improves operational efficiency and security. In order to meet different industrial needs, this project offers a multifaceted strategy that makes use of GPS, blockchain, transfer learning (TL), and convolutional neural networks (CNN). It offers services in data integrity assurance, supply chain management, continuous machine monitoring, and smooth data integration through Software Defined Networking (SDN). The total performance of the supply chain is enhanced by this system, which guarantees reliable asset tracking, effective machine maintenance, and safe data management. Result and Discussion Different classifiers' performance under diverse oversampling strategies yields a range of outcomes. DB-CGAN has the greatest F1-score for the Random Forest With an F1-score of 88.70%, the Deep Neural Network (DNN) classifier outperforms SMOTE (87%), MENGNETO (84%), and ADASYN (86%). With DB-CGAN at 92.72%, the Convolutional Neural Network (CNN) classifier performs remarkably well. SMOTE and ADASYN are next in line at 88%, and MENGNETO comes in at 89%. In conclusion, out of all the classifiers, DB-CGAN consistently produces the greatest F1-scores. Table 1 comparison of several classifiers using the F1 score with other networks Classifier SMOTE ADASYN MENGNETO DB-CGAN F 1 (%) F 1 (%) F 1 (%) F 1 (%) RF 86 89 88 90.21 SVM 88 84 88 91.32 DNN 87 84 86 88.70 CNN 88 89 88 92.72 Several algorithm types, including SMOTE, ADASYN, MENGNETO, and DB-CGAN, are compared with Random Forest (RF), Support Vector Machine (SVM), and Deep Neural Network (DNN). It is demonstrated that, for IIOT, the CNN-based DB-CGAN algorithms perform better than the other classifier types. Table 2 CNN in comparison to other neural networks Convolutional Neural Network Training time in seconds Accuracy SENet 1800 0.9 GoogleNet 330 0.899 ResNet-50 2100 0.92 Inception v2 2200 0.9 Inception v3 6500 0.9 RNN 1900 0.82 FNN 2000 0.8 RF 3500 0.8 SVM 7000 0.89 DNN 1000 0.8 The CNN-GoogleNet performs better with less training time and higher accuracy when compared to other types of convolutional neural networks. It has been demonstrated that the network with better accuracy and less training is efficient because as training time increases, the number of epochs increases as well, which causes the network to overtrain, where the neural networks attempt to memorize the values. Here, industry data is provided, including defect occurrence, operation time, client purchase, and others. CNN is used to test the efficiency, and the resulting output is taken into consideration. Additionally, based on accuracy, it is anticipated that the output would be provided accurately based on fault correction if any machines are repaired, for customer satisfaction, and other tasks completed flawlessly when utilizing CNN. Several methods, including SMOTE, ADASYN, MENGNETO, and DB-CGAN, were used to assess the classification performance across diverse models: F1 scores for Random Forest (RF) varied from 86–90.21%. F1 scores for the Support Vector Machine (SVM) varied from 84–91.32%. F1 scores for Deep Neural Networks (DNNs) varied from 84–88.70%. F1 scores for the Convolutional Neural Network (CNN) varied from 88–92.72%. The CNN-based DB-CGAN algorithm outperformed the others, proving that the suggested approach is effective in securely and precisely managing IIoT data. Neural Network Comparison Training time and accuracy were used to compare several Convolutional Neural Network architectures: SENet: 90% accuracy, 1800 seconds. GoogleNet: 89.9% accuracy, 330 seconds. ResNet-50: 92% accuracy, 2100 seconds. Inception v2: 90% accuracy, 2200 seconds. Inception v3: 90% accuracy, 6500 seconds. Recurrent Neural Network (RNN): 82% accuracy, 1900 seconds. Forward Neural Network (FNN): 80% accuracy, 2000 seconds. Random Forest (RF): 80% accuracy, 3500 seconds. Support Vector Machine (SVM): 89% accuracy, 7000 seconds. Deep Neural Network (DNN): 80% accuracy, 1000 seconds. The most effective network was GoogleNet, which balanced high accuracy with little training time. Because it avoids overtraining and guarantees that the network generalizes effectively to new data, its efficiency is essential. Graphical Analysis The CNN model's training accuracy and loss are shown in Figs. 2 and 3, which show little data loss and steadily rising accuracy. The usefulness of TL in real-time machine monitoring and maintenance is highlighted by Fig. 4, which contrasts it with other algorithms and shows its better true positive rate (97% accuracy). The effectiveness of CNN[ 16 ] and graph-based features in detecting and eliminating malware is demonstrated in Table 2 and Fig. 1 , and the results in Figs. 2 and 3 make it evident that data loss is extremely low and accuracy is increasing. Additionally, the use of SDN networks with IIoT guarantees dependability and faster communication [ 11 ][ 12 ]. Additionally, the idea of directly inserting data into blocks greatly aids in creating a cohesive product [ 13 ]. Once the data is compressed and saved, the client has additional possibilities. Furthermore, with 97% accuracy, Fig. 4 makes it abundantly evident that the TL performs better than the other algorithms. Conclusion By combining safe data management with real-time customer service, the "GPS-Enabled Secured Software Defined Network Using Transfer Learning and Blockchain Technology for Industrial IoT" tackles the technical backlog in industry. This method improves supply chain efficiency by using blockchain for data integrity, GPS for tracking, and transfer learning to adaptively detect problems and send out notifications. This project places a higher priority on data integrity than data science, which is crucial for major industries. In order to guarantee reliable operations and strong and secure data integration, blockchain technology is used in conjunction with SDN and CNN. By incorporating cutting-edge technologies like TL, CNN, and blockchain, the suggested methodology successfully improves IIoT applications. This integration greatly enhances supply chain management by guaranteeing safe, effective, and dependable industrial operations. The outcomes validate the robustness of the suggested method by showcasing the effectiveness of GoogleNet and the superiority of CNN-based DB-CGAN algorithms. In order to update industrial processes, this project offers a comprehensive solution that addresses data integrity, real-time monitoring, and effective communication. Declarations Author Contribution Only Author A wrote all the manuscript References Ailyn D, Gilbert J (2024) Internet of Things (IoT) and its Integration with Telecommunication Networks. Computer Science Challenges and Opportunities Akrasi-Mensah N, Kwadwo AS, Agbemenu H, Nunoo-Mensah ET, Tchao A-R, Ahmed (2022) Eliel Keelson, Axel Sikora, Dominik Welte, and Jerry John Kponyo. Adaptive Storage Optimization Scheme for Blockchain-IIoT Applications Using Deep Reinforcement Learning. IEEE Access Al-Balasmeh H, Singh M, Singh R (2024) Comprehensive Review of Location Privacy Preservation Techniques in Location-Based Services (LBS). Innovative and Intelligent Digital Technologies; Towards an Increased Efficiency: Volume 1, 691–705 Cui J, Liu N, Zhang Q, He D, Gu C, Zhong H (2022) Efficient and Anonymous Cross-Domain Authentication for IIoT Based on Blockchain. IEEE Trans Netw Sci Eng Gupta M, Jain K (2024) A Comprehensive Survey of Aerial Mesh Networks (AMN): Characteristics, Application, Open Issues, Challenges, and Research Directions. Wireless Pers Commun 138(1):333–368 Hindistan Y, Selim, Fatih Yetkin E (2023) A Hybrid Approach with GAN and DP for Privacy Preservation of IIoT Data. IEEE Access Ji L, He S, Wu W, Gu C, Bi J, Shi Z (2021) Dynamic Network Slicing Orchestration for Remote Adaptation and Configuration in Industrial IoT. IEEE Trans Industr Inf 18(6):4297–4307 Krishnan P, Jain K (2021) KrishnashreeAchuthan, and Rajkumar Buyya. Software-defined security-by-contract for blockchain-enabled MUD-aware Industrial IoT edge networks. IEEE Trans Industr Inf Nguyen TN, Ngo Q-D, Nguyen H-T, Nguyen LG (2022) An Advanced Computing Approach for IoT-Botnet Detection in Industrial Internet of Things. IEEE Trans Industr Inf Rathee G, Ahmad F, Jaglan N, CharalambosKonstantinou (2022) A Secure and Trusted Mechanism for Industrial IoT Network using Blockchain. arXiv preprint arXiv:220603419 Sharma S, Popli R, Singh S, Chhabra G, Saini GS, Singh M, Kumar R (2024) The Role of 6G Technologies in Advancing Smart City Applications: Opportunities and Challenges. Sustainability 16(16):7039 Singh A, Kumar K, Chatterjee, Singh A (2022) An image security model based on chaos and DNA cryptography for IIoT images. IEEE Trans Industr Inf 19(2):1957–1964 Soleymani SA, Goudarzi S, Anisi MH, Cruickshank H, Jindal A, Nazri Kama (2022) TRUTH: Trust and Authentication Scheme in 5G-IIoT. IEEE Trans Industr Inf 19(1):880–889 Song L, Sun G, Yu H, Niyato D (2025) ESPD-LP: Edge Service Pre-Deployment Based on Location Prediction in MEC. IEEE Transactions on Mobile Computing Tariq U, Tariq B (2025) Signal Characteristic Analysis and Anomaly Detection for GPS Spoofing Mitigation. Ubiquitous Technol J 1(1):10–22 Thilakarathne NN, Bakar MSA, Abas PE, Yassin H (2025) Internet of Things Enabled Smart Agriculture: Current Status, Latest Advancements, Challenges and Countermeasures. Heliyon Ullah FU, Min K, Muhammad I, UlHaq N, Khan AA, Heidari S, WookBaik, and Victor, Hugo C, de Albuquerque (2021) AI-Assisted Edge Vision for Violence Detection in IoT-Based Industrial Surveillance Networks. IEEE Transactions on Industrial Informatics 18, no. 8 : 5359–5370 Wang T, Fang K, Wei W, Tian J, Pan Y, Li J (2022) Microcontroller Unit Chip Temperature Fingerprint Informed Machine Learning for IIoT Intrusion Detection. IEEE Trans Industr Inf 19(2):2219–2227 Wu H, Chen J, Nguyen TN (2022) Lyapunov-Guided Delay-Aware Energy Efficient Offloading in IIoT-MEC Systems. IEEE Trans Industr Inf 19(2):2117–2128 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5963243","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":411887257,"identity":"56ee11e4-52a9-4f36-ae77-8dcc37eef894","order_by":0,"name":"BAZEER AHAMED BAGRUDEEN","email":"data:image/png;base64,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","orcid":"","institution":"UNIVERSITY OF TECHNOLOGY AND APPLIED SCIENCES","correspondingAuthor":true,"prefix":"","firstName":"BAZEER","middleName":"AHAMED","lastName":"BAGRUDEEN","suffix":""}],"badges":[],"createdAt":"2025-02-05 07:23:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5963243/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5963243/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":75701841,"identity":"db467626-6230-4637-96e0-4003dc62fd4a","added_by":"auto","created_at":"2025-02-07 09:35:00","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":41033,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSurveillance and Maintenance Service Provided by the IIOT\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5963243/v1/5f2774505799fac1d28d83de.png"},{"id":75702950,"identity":"d3d42fef-70c7-4785-99c9-b2d80b721a49","added_by":"auto","created_at":"2025-02-07 09:43:00","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":122841,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGraph showing the training accuracy\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5963243/v1/85aafe376f7bab5b530ecdd5.png"},{"id":75701844,"identity":"15d50222-1914-4972-ba55-4502165fdd01","added_by":"auto","created_at":"2025-02-07 09:35:00","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":182943,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGraph showing the training loss\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5963243/v1/b1f88068073ceb217413e664.png"},{"id":75702952,"identity":"2643a9c4-9cec-4af8-8c36-3acead4a06b7","added_by":"auto","created_at":"2025-02-07 09:43:01","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":147003,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of TL with other algorithms for true positive rate Vs False Positive Rate\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5963243/v1/96c955e22a8431f23549c9c1.png"},{"id":93410948,"identity":"d29e7c98-272f-496c-b1e3-6801f4934500","added_by":"auto","created_at":"2025-10-13 14:32:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":971447,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5963243/v1/00fc4c72-6371-4b95-b62d-8fb604264170.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"GSTLB: A Secure GPS-Enabled Software-Defined Network Leveraging Transfer Learning and Blockchain for Industrial IoT (IIoT)","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSmart automation's incorporation of the Industrial Internet of Things (IIoT) closes the gap between cutting-edge technology and conventional industry methods. In order to improve communication, surveillance, and operational efficiency, this project uses 5G networks, transfer learning (TL), and blockchain technology to present a novel approach to asset tracking and machine problem identification using GPS and sophisticated analytics. Customers can keep an eye on products all the way through the supply chain with real-time GPS monitoring, and sensors can identify machine problems and send out maintenance notifications to save downtime. 5G's fast speeds facilitate data analysis and real-time monitoring. Blockchain technology guards against malicious data insertion and eavesdropping, protecting data from GPS monitoring, surveillance, and alarms. Plot-based feature extraction and Convolutional Neural Network (CNN)-based feature extraction are combined in a hybrid security approach to better separate harmful from valid data, hence improving security. Software-Defined Networking (SDN) offers centralized network control, while network slicing maximizes resource allocation. The blockchain compresses data to make it easier to store and retrieve. According to simulation results, this approach greatly improves IIoT network security, permits faster data transfer, and makes it easier to conduct ongoing machine and product inspections, all of which increase supply chain efficiency. A more effective and safe industrial ecosystem is made possible by this all-encompassing approach, which modernizes industrial operations and guarantees improved asset tracking, machine maintenance, and data protection.\u003c/p\u003e"},{"header":"Literature review","content":"\u003cp\u003eIIoT focuses on monitoring the PH level [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e] and temperature of the contaminated water and when it encounters the water flows with the highest contamination level, the water flow valves were controlled and that water was sent to the water treatment plant and processed and sent to the plant and these data such as monitoring of PH evel and temperature of water by sensor, excess flow of waste water everything is monitored. The data is collected utilising blockchain [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e], where the blockchain technology aids in the expansion of the supply chain.H1: Ownership concentrationhas no effect on firm-level stock return.\u003c/p\u003e \u003cp\u003eThey suggested a heuristic algorithm using AI and the idea of network slicing to replace the old way of utilising the physical network, and they created the concept of network slicing orchestration system[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], so that various applications may utilise their specialised network. In IIOT, all of these processes are active at the same time.\u003c/p\u003e \u003cp\u003eWhen considering the of IOT Botnet in IIOT[4,] for trying to overcome an approach called graph based feature which can help in differentiating or differentiating both malicious and benign samples, for gathering data samples among the 3 techniques available in botnet detection, dynamic analysis methods and produces an accuracy of 98.1% and 91.99% in botnet detection\u003c/p\u003e \u003cp\u003eSurveillance [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] has grown increasingly important in recent years, yet cameras can still identify people and objects. To boost trust in IIOT-based monitoring, AI has been applied, which aids in the detection of dangerous objects, as well as the usage of a violence detection network. After the harmful object is recognised, an alert is sent to the appropriate departments, and CNN is used for frame identification.\u003c/p\u003e \u003cp\u003eBecause of numerous entities such as packaging units, hubs, and manufacturing units, IIOT substantially lags in data integration, therefore data is merged via block chain technology [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], where data transparency is assured safely and hostile device attacks are avoided\u003c/p\u003e \u003cp\u003eThe technological advances in IIoT [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] lead to the development of light weight devices with less energy consumption and longer range of wireless connection, because of this there has been a development such as smart airport, where enormous sensors has incorporated and enabled various wireless technologies such as Wi-Fi, and so on, which enable data collection and adaptation of system to the varying environment. The use of Industrial Internet of Things (IIoT) in smart airports has brought about numerous benefits such as improved communication, business processes, and efficiency. However, it has also introduced new vulnerabilities that can be exploited by cyber attackers to compromise both the digital infrastructure and physical assets of the airport. To address this issue, a new security-oriented IIoT testbed called SAir-IIoT has been developed. This testbed comprises multiple IIoT devices and communication protocols that are organized into distinct zones and can be remotely accessed as a service. The purpose of this paper is to present SAir-IIoT as a novel solution to the security challenges posed by IIoT in smart airports. Our comparison of SAir-IIoT with other IIoT-based testbeds reveals its complexity and effectiveness in evaluating new cyber security methods. SAir-IIoT is compared with other IIoT-based testbeds in order to demonstrate its complexity and effectiveness when evaluating new cyber security methods. In conclusion, we compare SAir-IIoT with other IIoT-based testbeds, revealing both its complexity and effectiveness. As a final result, we compare SAir-IIoT with other IIoT-based testbeds, demonstrating the complexity and effectiveness of this testbed with respect to evaluating new cyber security approaches. Our final analysis compares SAir-IIoT with other IIoT-based testbeds in order to reveal its complexity and suitability for the evaluation of new cyber security methods. Our final evaluation of SAir-IIoT reveals its complexity and effectiveness in evaluating new cyber security techniques. In addition, we compare SAir-IIoT to other IIoT-based testbeds to demonstrate its versatility and effectiveness in assessing new cybersecurity techniques. We finalize our study by comparing SAir-IIoT with other IIoT-based testbeds, in order to demonstrate the complexity and effectiveness of the testbed when it comes to evaluating new cyber security methods. In addition, we have compared SAir-IIoT with other IIoT-based testbeds, indicating its complexity and efficacy as a tool for testing new cyber security methodologies. Finally, we assess SAir-IIoT compared with other IIoT-based testbeds, revealing its complexity and effectiveness for the evaluation of new cyber security technologies.\u003c/p\u003e \u003cp\u003eIt is difficult to address the combined optimization problem of a 5G-inspired IIoT-MEC interactive network, which aims to maximize MNO income and minimize IIoT operators' economic cost. In order to increase MNO income while maintaining reasonable service charges for IIoT mobile devices (MDs), the author [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] of this letter suggests a dynamic pricing model for IIoT-MEC networks. A discrete finite Markov decision process (MDP) is used as the initial model for the dynamic pricing issue. After that, this problem is resolved using the Q-learning technique. The findings demonstrate that the suggested dynamic pricing method may greatly increase MNOs' income and lower the economic cost of IIoT operators.\u003c/p\u003e \u003cp\u003eThe paper [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] that proposes a novel method for preserving sensitive data in IIoT operations. The Industrial Internet of Things (IIoT) is being increasingly used in manufacturing and related industries, and Machine Learning (ML) techniques are commonly used to analyze the collected data. However, ML applications require decrypted data to perform efficiently, which raises concerns about data privacy and security. To address these concerns, the paper proposes a hybrid method of Generative Adversarial Networks (GAN) and Differential Privacy (DP) to preserve sensitive data in IIoT operations. The proposed method aims to sustain IIoT data privacy with minimal accuracy loss and additional computational costs. The paper uses publicly available datasets and a realistic IIoT dataset collected from a confectionery production process to demonstrate the efficiency of the proposed approach. The paper employs privacy assessment metrics and shows that the proposed method preserves the privacy of the data while maintaining the accuracy of the algorithms used for analysis. Overall, the method aims to prevent the production of hidden sensitive data from the sub-feature sets and ensure privacy protection for hidden sensitive data in IIoT operations.\u003c/p\u003e \u003cp\u003eModern urbanization [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] and smart cities have undergone a transformation thanks to recent advances in the Industrial Internet of Things (IIoT). Even though IIoT data contains a wealth of interesting events and objects, analyzing a sizable volume of IIoT data and creating predictions in real time are difficult. Recent developments in AI enable analysis of this enormous volume of IIoT data and the generation of insights for subsequent decision-making processes. In this post, we outline some essential features of IIoT data with AI support for monitoring smart cities. In order to gather events and objects from IIoT data in real time, we first integrated an AI-enabled IIoT framework with a crowdsourcing application that supports human intelligence. In order to automatically classify the collected events and objects and provide analytics, reports, and warnings from the IIoT data in real time, we have merged numerous AI algorithms that can operate on distributed edge and cloud nodes. Two scenarios can be used the results in. In the first instance, the smart city authority can validate the events that were processed by AI and allocate them to the proper authority for event management. In the second scenario, AI algorithms are permitted to communicate with people or the IIoT to complete tasks. The implementation specifics for the aforementioned situations as well as the test findings will be presented in the final section. The framework has the potential to be used in a smart city, according to the test findings.\u003c/p\u003e \u003cp\u003eDue to the dynamic and open nature of the Industrial Internet of Things (IIoT) [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] ecosystem, pictures produced by smart cameras and sensors are significantly at risk when sent over a public network. A practical strategy for protecting IIoT digital photos is encryption. This article discusses a framework for picture security that makes use of DNA cryptography and chaotic maps. The suggested approach generates three keys using a multilayer combination of a tent, circle, Chebyshev, and 3-D logistic map. These keys are utilized to build a key image on which DNA XOR operation is carried out to obtain the encrypted image, decide the rule to do DNA encoding-decoding on the subblocks, and perform row-column rotation of the subblocks. The result analysis of the suggested scheme shows that its average NPCR (99.6566%), UACI (33.4588%), Entropy (7.9971), and bigger key-space of 10195 are superior to those of the existing schemes and more resilient to various assaults.\u003c/p\u003e \u003cp\u003eIncreased security, transparency, and traceability are promised as a result of integrating [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] blockchain-IIoT into industrial operations. This development, however, runs into serious storage and scalability problems with current blockchain systems. A complete copy of the ledger is kept by each peer in the blockchain network, and it is updated by consensus. This complete replication strategy puts a pressure on the peers' storage space and would rapidly exceed the storage capacity of IIoT devices with limited resources. In the literature, a number of methods utilizing compression, summarization, or various storing systems have been suggested. In the last several years, there has been a lot of research done on the utilization of cloud resources for blockchain storage. Block selection is still a significant issue with cloud resources and blockchain integration, though. In order to solve the block selection problem, which entails choosing the blocks to be moved to the cloud, this research suggests a deep reinforcement learning (DRL) method. By transforming the multi-objective optimization of block selection into a Markov decision process (MDP), we suggest a DRL technique to tackle our problem. We create a mock blockchain environment to practise and test our suggested DRL methodology. The block selection issue is solved using two DRL algorithms, Advantage Actor-Critic (A2C) and Proximal Policy Optimization (PPO), and their performance benefits are examined. In comparison to the complete replication strategy used by traditional blockchain systems, PPO and A2C achieve storage reductions on the blockchain peer of 47.8% and 42.9%, respectively. The run-time difference between the slowest DRL algorithm, A2C, and the benchmark evolutionary algorithms employed in prior publications is only 7.2 times, validating the benefits of the DRL algorithms. The outcomes of the simulation further demonstrate that our DRL algorithms offer a flexible and dynamic response to the time-critical blockchain-IIoT context.\u003c/p\u003e \u003cp\u003eThe industrial Internet of Things (IIoT) [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] network places a high value on data, therefore trust and data security are two of the main issues. In order to safeguard data transmitted via wireless communication, contend with unauthorized entities, and guarantee data integrity, we create a cloud-integrated 5G-IIoT network architecture supported by a three-party authenticated key exchange (AKE) protocol with privacy-preserving. In addition, we create a trust model based on the Dempster-Shafer theory to evaluate the reliability of information gathered by smart devices/sensor nodes. Our scheme has withstood several well-known assaults, according to security assessments, in the IIoT environment. Using a tool for automated validation of internet security protocols and apps, we also examined the correctness of our plan. Additionally, the experimental and performance assessment findings demonstrate that the suggested strategy outperforms other works in terms of accuracy, latency, trust, and throughput.\u003c/p\u003e \u003cp\u003eThis article discusses the challenges associated with handling delay-sensitive [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] and compute-intensive workloads in the context of Industrial Internet of Things (IIoT) systems, and proposes a solution based on mobile edge computing (MEC) to offload these tasks for processing. However, offloading tasks can lead to higher energy consumption, which can impact battery life, and the communication delay between IIoT devices and MEC can also impact performance. To address these issues, the article proposes a novel delay-aware energy-efficient (DAEE) online offloading algorithm that can adaptively offload more tasks when the network quality is good, while delaying transmission when connectivity is poor but ensuring that task deadlines are not violated. The article also discusses the theoretical underpinnings of the proposed algorithm and provides simulation results to demonstrate its effectiveness in minimizing energy consumption and maintaining low latency, especially for delay-sensitive and compute-intensive tasks.\u003c/p\u003e \u003cp\u003eThis article [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] discusses the security and privacy issues that arise when devices from different management domains in the Industrial Internet of Things (IIoT) communicate with each other. Existing authentication schemes have limitations such as a single point of failure in a trusted center, high certificate management costs, and low authentication efficiency. To address these issues, the article proposes an efficient and anonymous cross-domain authentication scheme based on blockchain technology. The scheme uses dynamic accumulator technology and combines it with blockchain to achieve fast authentication while ensuring device anonymity to prevent link ability of identities. The article also discusses the security analysis and performance evaluation of the proposed scheme, which demonstrates its ability to resist common attacks and its feasibility and efficiency in practice. Overall, the article proposes a novel approach to secure cross-domain communication in the IIoT using blockchain technology.\u003c/p\u003e \u003cp\u003eThis paper [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] discusses the challenges of extracting value from the vast amounts of data produced by Industrial Internet of Things (IIoT) infrastructure, which is increasingly integrated with artificial intelligence (AI) and machine learning (ML) solutions. Although digital marketplaces have emerged to enable data owners to monetize their data, concerns such as privacy, security, and fair payment settlement have hindered their adoption. Additionally, the current centralized platforms are controlled by large multinational corporations and lack transparency between buyers and sellers. To address these challenges, the article proposes a decentralized platform for a digital data marketplace for IoT data that leverages a decentralized data streaming network for reliable and fault-tolerant hosting of IoT data. The platform also ensures fair trading, data storage, and delivery in a privacy-preserving manner, and calculates trust metrics for actors in the network. The feasibility of the proposed platform is studied through the development of an open-source library using Hyperledger Fabric and VerneMQ, which is tested on a real-time Google cloud platform for throughput, overheads, and scalability. Overall, the article proposes a novel approach to enable fair and transparent monetization of IoT data in a decentralized manner.\u003c/p\u003e \u003cp\u003eThis paper discusses [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] the importance of physics-informed learning for safety in Industrial Internet of Things (IIoT) systems and proposes a new method for IIoT intrusion detection called MTID. The conventional intrusion detection methods require auxiliary equipment and are not versatile enough for general IIoT systems. MTID is based on the temperature fingerprint of the microcontroller unit (MCU) chip and uses a self-encoder-based intrusion detection model to identify the security status of the nodes. The temperature residuals dataset is constructed by analyzing the relationship between the temperature sequence and the computational complexity of the node. The proposed method is designed to be applicable under the diversified deployment environment of IIoT systems, and an online incremental training method is developed and applied to ensure the model's applicability. Experimental analysis using Raspberry Pi 4B shows that MTID achieves an accuracy of 89% for intrusion detection, which demonstrates the feasibility of the intrusion detection method based on MCU temperature.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Proposed Methodology","content":"\u003cp\u003eA number of cutting-edge technologies are integrated into the suggested methodology for the \"GPS Enabled Secured Software Defined Network using Transfer Learning and Blockchain Technology for Industrial IoT (IIoT)\" in order to accomplish safe, effective, and dependable industrial operations. GPS technology makes it possible to track machines and assets in real time, offering location-based services and ongoing monitoring. By applying transfer learning, pre-trained models are modified to fit the unique IIoT context, improving the capacity to recognize machine problems and efficiently initiate maintenance alerts. By safely storing GPS tracking, surveillance, and alert data and guarding against eavesdropping and malicious data insertion, blockchain technology guarantees data integrity. A hybrid security approach uses both plot-based and CNN-based feature extraction to differentiate between malicious and legitimate data. Network slicing optimizes resource allocation by establishing virtual networks customized to particular applications and requirements, while Software Defined Networking (SDN) offers centralized network control, facilitating effective administration and integration of various IIoT data streams.\u003c/p\u003e\u003cp\u003eFour types of services are provided by the suggested method: The supplier can ensure payment, storage, quality control, and ordering. The worker gains knowledge in planning, reporting, human resources, and goal attainment. Order planning, logistics, and services all benefit customers. Additionally, machines can be programmed to understand uptime and downtime.\u003c/p\u003e\u003cp\u003eBlockchain technology compresses data for effective storage and retrieval, and unified product data management improves customer choices and data management in general. DB-CGAN algorithms are used in conjunction with CNNs to improve classification performance, and CNN-based feature extraction is used for real-time data monitoring to identify and eliminate fraudulent inputs, guaranteeing data integrity. According to simulation results, the suggested methodology greatly improves supply chain efficiency by enabling faster data transfer, enhancing IIoT network security, and facilitating ongoing product and machine inspection. Using algorithms like SMOTE, ADASYN, MENGNETO, and DB-CGAN, several classifiers (RF, SVM, DNN, and CNN) are compared; CNN-based DB-CGAN yields the best results. Additionally, various CNN architectures (e.g., GoogleNet, ResNet-50) are assessed for accuracy and training time; GoogleNet is found to be the most effective. The suggested solution seeks to improve the overall security, effectiveness, and dependability of IIoT applications by integrating these techniques, guaranteeing reliable asset tracking, effective machine maintenance, and safe data handling.\u003c/p\u003e\u003cp\u003eContinuous machine monitoring, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e equipment or products, is the second type of service. The TL[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] uses its experience to determine when the downtime and uptime will occur; if there is a deviation, the maintenance manager will be informed and the necessary action will be taken. Additionally, there is a chance that malicious data may be added unintentionally or intentionally when using IIoT. For this reason, data will be continuously monitored by a Convolutional Neural Network and a graph feature-based method. If any deviations are discovered, the data is eliminated and the network is restored to normal., This is the third type of service and the fourth is SDN [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] service; however, because of the numerous components in the industry, including hubs, production, and packaging, data integration becomes difficult because of network slicing. The block chain is used to combine all of these data, and the SDN service keeps an eye on them for optimal security. The supply chain's growth is facilitated by this data integration.\u003c/p\u003e\u003cp\u003eThrough sophisticated data management and real-time monitoring, the incorporation of Industrial Internet of Things (IIoT) technologies into industrial processes improves operational efficiency and security. In order to meet different industrial needs, this project offers a multifaceted strategy that makes use of GPS, blockchain, transfer learning (TL), and convolutional neural networks (CNN). It offers services in data integrity assurance, supply chain management, continuous machine monitoring, and smooth data integration through Software Defined Networking (SDN). The total performance of the supply chain is enhanced by this system, which guarantees reliable asset tracking, effective machine maintenance, and safe data management.\u003c/p\u003e"},{"header":"Result and Discussion","content":"\u003cp\u003eDifferent classifiers' performance under diverse oversampling strategies yields a range of outcomes. DB-CGAN has the greatest F1-score for the Random Forest With an F1-score of 88.70%, the Deep Neural Network (DNN) classifier outperforms SMOTE (87%), MENGNETO (84%), and ADASYN (86%). With DB-CGAN at 92.72%, the Convolutional Neural Network (CNN) classifier performs remarkably well. SMOTE and ADASYN are next in line at 88%, and MENGNETO comes in at 89%. In conclusion, out of all the classifiers, DB-CGAN consistently produces the greatest F1-scores.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ecomparison of several classifiers using the F1 score with other networks\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClassifier\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSMOTE\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eADASYN\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMENGNETO\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDB-CGAN\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eF\u003csub\u003e1\u003c/sub\u003e(%)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eF\u003csub\u003e1\u003c/sub\u003e(%)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eF\u003csub\u003e1\u003c/sub\u003e(%)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eF\u003csub\u003e1\u003c/sub\u003e(%)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRF\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e86\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e89\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e88\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e90.21\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSVM\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e88\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e88\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e91.32\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDNN\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e87\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e86\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e88.70\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCNN\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e88\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e89\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e88\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e92.72\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003eSeveral algorithm types, including SMOTE, ADASYN, MENGNETO, and DB-CGAN, are compared with Random Forest (RF), Support Vector Machine (SVM), and Deep Neural Network (DNN). It is demonstrated that, for IIOT, the CNN-based DB-CGAN algorithms perform better than the other classifier types.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCNN in comparison to other neural networks\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConvolutional Neural Network\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining time in seconds\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSENet\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1800\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoogleNet\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e330\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.899\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResNet-50\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2100\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInception v2\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2200\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInception v3\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6500\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRNN\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1900\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFNN\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2000\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3500\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSVM\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7000\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDNN\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1000\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003eThe CNN-GoogleNet performs better with less training time and higher accuracy when compared to other types of convolutional neural networks. It has been demonstrated that the network with better accuracy and less training is efficient because as training time increases, the number of epochs increases as well, which causes the network to overtrain, where the neural networks attempt to memorize the values. Here, industry data is provided, including defect occurrence, operation time, client purchase, and others. CNN is used to test the efficiency, and the resulting output is taken into consideration. Additionally, based on accuracy, it is anticipated that the output would be provided accurately based on fault correction if any machines are repaired, for customer satisfaction, and other tasks completed flawlessly when utilizing CNN.\u003c/p\u003e\u003cp\u003eSeveral methods, including SMOTE, ADASYN, MENGNETO, and DB-CGAN, were used to assess the classification performance across diverse models:\u003c/p\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eF1 scores for Random Forest (RF) varied from 86–90.21%.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eF1 scores for the Support Vector Machine (SVM) varied from 84–91.32%.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eF1 scores for Deep Neural Networks (DNNs) varied from 84–88.70%.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eF1 scores for the Convolutional Neural Network (CNN) varied from 88–92.72%.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003eThe CNN-based DB-CGAN algorithm outperformed the others, proving that the suggested approach is effective in securely and precisely managing IIoT data.\u003c/p\u003e\n\u003ch3\u003eNeural Network Comparison\u003c/h3\u003e\n\u003cp\u003eTraining time and accuracy were used to compare several Convolutional Neural Network architectures:\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\u003cul\u003e \u003cli\u003e \u003cp\u003eSENet: 90% accuracy, 1800 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eGoogleNet: 89.9% accuracy, 330 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eResNet-50: 92% accuracy, 2100 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eInception v2: 90% accuracy, 2200 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eInception v3: 90% accuracy, 6500 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eRecurrent Neural Network (RNN): 82% accuracy, 1900 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eForward Neural Network (FNN): 80% accuracy, 2000 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eRandom Forest (RF): 80% accuracy, 3500 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eSupport Vector Machine (SVM): 89% accuracy, 7000 seconds.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eDeep Neural Network (DNN): 80% accuracy, 1000 seconds.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003eThe most effective network was GoogleNet, which balanced high accuracy with little training time. Because it avoids overtraining and guarantees that the network generalizes effectively to new data, its efficiency is essential.\u003c/p\u003e\n\u003ch3\u003eGraphical Analysis\u003c/h3\u003e\n\u003cp\u003eThe CNN model's training accuracy and loss are shown in Figs.\u0026nbsp;2 and 3, which show little data loss and steadily rising accuracy. The usefulness of TL in real-time machine monitoring and maintenance is highlighted by Fig.\u0026nbsp;4, which contrasts it with other algorithms and shows its better true positive rate (97% accuracy).\u003c/p\u003e \u003cp\u003eThe effectiveness of CNN[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] and graph-based features in detecting and eliminating malware is demonstrated in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, and the results in Figs.\u0026nbsp;2 and 3 make it evident that data loss is extremely low and accuracy is increasing. Additionally, the use of SDN networks with IIoT guarantees dependability and faster communication [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e][\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Additionally, the idea of directly inserting data into blocks greatly aids in creating a cohesive product [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Once the data is compressed and saved, the client has additional possibilities. Furthermore, with 97% accuracy, Fig.\u0026nbsp;4 makes it abundantly evident that the TL performs better than the other algorithms.\u003c/p\u003e "},{"header":"Conclusion","content":"\u003cp\u003eBy combining safe data management with real-time customer service, the \"GPS-Enabled Secured Software Defined Network Using Transfer Learning and Blockchain Technology for Industrial IoT\" tackles the technical backlog in industry. This method improves supply chain efficiency by using blockchain for data integrity, GPS for tracking, and transfer learning to adaptively detect problems and send out notifications. This project places a higher priority on data integrity than data science, which is crucial for major industries. In order to guarantee reliable operations and strong and secure data integration, blockchain technology is used in conjunction with SDN and CNN. By incorporating cutting-edge technologies like TL, CNN, and blockchain, the suggested methodology successfully improves IIoT applications. This integration greatly enhances supply chain management by guaranteeing safe, effective, and dependable industrial operations. The outcomes validate the robustness of the suggested method by showcasing the effectiveness of GoogleNet and the superiority of CNN-based DB-CGAN algorithms. In order to update industrial processes, this project offers a comprehensive solution that addresses data integrity, real-time monitoring, and effective communication.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eOnly Author A wrote all the manuscript\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAilyn D, Gilbert J (2024) Internet of Things (IoT) and its Integration with Telecommunication Networks. Computer Science Challenges and Opportunities\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAkrasi-Mensah N, Kwadwo AS, Agbemenu H, Nunoo-Mensah ET, Tchao A-R, Ahmed (2022) Eliel Keelson, Axel Sikora, Dominik Welte, and Jerry John Kponyo. Adaptive Storage Optimization Scheme for Blockchain-IIoT Applications Using Deep Reinforcement Learning. IEEE Access\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAl-Balasmeh H, Singh M, Singh R (2024) Comprehensive Review of Location Privacy Preservation Techniques in Location-Based Services (LBS). Innovative and Intelligent Digital Technologies; Towards an Increased Efficiency: Volume 1, 691\u0026ndash;705\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui J, Liu N, Zhang Q, He D, Gu C, Zhong H (2022) Efficient and Anonymous Cross-Domain Authentication for IIoT Based on Blockchain. IEEE Trans Netw Sci Eng\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGupta M, Jain K (2024) A Comprehensive Survey of Aerial Mesh Networks (AMN): Characteristics, Application, Open Issues, Challenges, and Research Directions. Wireless Pers Commun 138(1):333\u0026ndash;368\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHindistan Y, Selim, Fatih Yetkin E (2023) A Hybrid Approach with GAN and DP for Privacy Preservation of IIoT Data. IEEE Access\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJi L, He S, Wu W, Gu C, Bi J, Shi Z (2021) Dynamic Network Slicing Orchestration for Remote Adaptation and Configuration in Industrial IoT. IEEE Trans Industr Inf 18(6):4297\u0026ndash;4307\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKrishnan P, Jain K (2021) KrishnashreeAchuthan, and Rajkumar Buyya. Software-defined security-by-contract for blockchain-enabled MUD-aware Industrial IoT edge networks. IEEE Trans Industr Inf\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen TN, Ngo Q-D, Nguyen H-T, Nguyen LG (2022) An Advanced Computing Approach for IoT-Botnet Detection in Industrial Internet of Things. IEEE Trans Industr Inf\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRathee G, Ahmad F, Jaglan N, CharalambosKonstantinou (2022) A Secure and Trusted Mechanism for Industrial IoT Network using Blockchain. arXiv preprint arXiv:220603419\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma S, Popli R, Singh S, Chhabra G, Saini GS, Singh M, Kumar R (2024) The Role of 6G Technologies in Advancing Smart City Applications: Opportunities and Challenges. Sustainability 16(16):7039\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh A, Kumar K, Chatterjee, Singh A (2022) An image security model based on chaos and DNA cryptography for IIoT images. IEEE Trans Industr Inf 19(2):1957\u0026ndash;1964\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoleymani SA, Goudarzi S, Anisi MH, Cruickshank H, Jindal A, Nazri Kama (2022) TRUTH: Trust and Authentication Scheme in 5G-IIoT. IEEE Trans Industr Inf 19(1):880\u0026ndash;889\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong L, Sun G, Yu H, Niyato D (2025) ESPD-LP: Edge Service Pre-Deployment Based on Location Prediction in MEC. IEEE Transactions on Mobile Computing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTariq U, Tariq B (2025) Signal Characteristic Analysis and Anomaly Detection for GPS Spoofing Mitigation. Ubiquitous Technol J 1(1):10\u0026ndash;22\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThilakarathne NN, Bakar MSA, Abas PE, Yassin H (2025) Internet of Things Enabled Smart Agriculture: Current Status, Latest Advancements, Challenges and Countermeasures. Heliyon\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUllah FU, Min K, Muhammad I, UlHaq N, Khan AA, Heidari S, WookBaik, and Victor, Hugo C, de Albuquerque (2021) AI-Assisted Edge Vision for Violence Detection in IoT-Based Industrial Surveillance Networks. IEEE Transactions on Industrial Informatics 18, no. 8 : 5359\u0026ndash;5370\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang T, Fang K, Wei W, Tian J, Pan Y, Li J (2022) Microcontroller Unit Chip Temperature Fingerprint Informed Machine Learning for IIoT Intrusion Detection. IEEE Trans Industr Inf 19(2):2219\u0026ndash;2227\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu H, Chen J, Nguyen TN (2022) Lyapunov-Guided Delay-Aware Energy Efficient Offloading in IIoT-MEC Systems. IEEE Trans Industr Inf 19(2):2117\u0026ndash;2128\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"SDN, Block Chain Technology, Transfer Learning, IIoT, network security","lastPublishedDoi":"10.21203/rs.3.rs-5963243/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5963243/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIIoT sensors bridge the gap between industry and technology by assisting in intelligent automation of the sector. The proposed method uses GPS to track assets and identify products for customers. It also uses TL (Transfer Learning) to identify machine problems, which alerts for immediate service. Using the 5G network greatly facilitates faster communication and surveillance control. All of the data from these various applications, including GPS location tracking, surveillance, and alerting, are stored in a block chain where malicious data can be added with the original data or eves dropping is possible, which could interfere with the industry's normal operations. Plot-based features and CNN-based feature extraction are both used to get around this, and the network's efficiency and security are increased by this hybrid approach of separating dangerous input from the original data. When using the block chain methodology, data is compressed to make it unified because Network Slicing is the primary concept utilized in industry and SDN (Software Defined Network) is used as a centralized one for communication. According to the simulation results, the suggested methodology improves IIoT network security, speeds up data transmission, and conducts ongoing product and machine inspections, all of which significantly aid in expanding the supply chain.\u003c/p\u003e","manuscriptTitle":"GSTLB: A Secure GPS-Enabled Software-Defined Network Leveraging Transfer Learning and Blockchain for Industrial IoT (IIoT)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-07 09:34:56","doi":"10.21203/rs.3.rs-5963243/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"42bfce8f-10c7-4f8d-a445-0018dfcc2911","owner":[],"postedDate":"February 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-13T14:24:02+00:00","versionOfRecord":[],"versionCreatedAt":"2025-02-07 09:34:56","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5963243","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5963243","identity":"rs-5963243","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}