The rapid growth of data generated from digital platforms, sensors, and connected devices has accelerated the integration of Artificial Intelligence (AI) with Big Data technologies. AI-driven Big Data systems enable advanced data processing, predictive analytics, and intelligent decision-making across diverse application domains. This review paper provides a comprehensive overview of the current trends, key challenges, and future research directions in AI-driven Big Data technologies. It discusses emerging trends such as automated analytics, deep learning–based data modeling, real-time processing, and the convergence of AI with cloud, edge, and Internet of Things (IoT) environments. The paper also highlights critical challenges, including data privacy and security concerns, scalability issues, high computational costs, and data quality and bias. Furthermore, it explores promising future directions, such as explainable and trustworthy AI, hybrid intelligence models, and sustainable AI-driven data ecosystems. This review aims to offer valuable insights for researchers and practitioners by summarizing recent advancements and identifying open research opportunities in the evolving landscape of AI-driven Big Data technologies

This research, guided by the Conceive-Design-Implement-Operate (CDIO) model, emphasizes full-cycle teaching reform and practice within the context of an Internet of Things (IoT) cloud platform application development. The primary focus is on enhancing students' engineering practice capabilities, deeply exploring and implementing the entire "ideation-design-implementation-operation" process in effective teaching methodologies. Through this research, our goal is to fully incorporate CDIO's project-driven concept into professional IoT teaching. This approach enables students to master a comprehensive skill set, ranging from demand analysis to system operation and maintenance, within real project scenarios. Consequently, it enhances their ability to tackle complex engineering problems.

The rapid convergence of Fog, Cloud, and Internet of Things (IoT) technologies has introduced a new era of distributed intelligence and real-time data processing. However, ensuring secure, reliable, and energy-efficient communication across heterogeneous and resource-constrained nodes remains a fundamental challenge. This paper introduces a novel framework entitled Federated Learning-Based Trust and Energy-Aware Routing (FL-TEAR), designed to enhance routing performance in hybrid Fog-Cloud-IoT environments through collaborative intelligence, adaptive trust management, and dynamic energy optimization. The FL-TEAR system replaces static trust evaluation with a federated learning paradigm, allowing IoT and fog nodes to cooperatively train a global trust-energy model without exposing raw data. Trust scores are continuously refined based on behavioral patterns, communication reliability, and residual energy, while routing paths are selected using a composite fitness function integrating trustworthiness, energy availability, latency, and link stability. The hierarchical architecture, spanning IoT, fog, and cloud layers, reduces communication overhead, supports scalability, and preserves privacy. Simulation results confirm that FL-TEAR significantly outperforms state-of-the-art baselines such as E-ODMA (Energy-Efficient On-Demand Multipath Adaptive) + AOMDV (Ad hoc On-Demand Multipath Distance Vector), TAGA (Trust-Aware Geographic Routing Algorithm), and EigenTrust, achieving approximately 23% higher trust accuracy, 23% lower energy consumption, approximately 13% greater packet delivery ratio, and 37% lower delay. These findings demonstrate that federated learning can effectively balance security, sustainability, and quality of service (QoS) in large-scale IoT ecosystems, establishing FL-TEAR as a viable pathway toward intelligent, secure, and energy-efficient next-generation networks.

The products derived from aquatic sources are nutritious due to their abundance of moisture, protein, and various other nutrients. Nonetheless, their vulnerability to deterioration and spoilage is significant when affected by microorganisms and internal enzymes. In order to extend the storage period and shelf life of aquatic products, we have developed a gas control system for preserving seafood with modified atmosphere packaging. At its core, this system integrates technologies from the Internet of Things, sensors, network communication, and PLC control systems. This system connects the PLC controller, the Internet of Things cloud box, and the remote client to ensure uninterrupted communication. Utilizing the TCP/IP communication protocol and 4G network, the Internet of Things cloud box acts as a data transmission transceiver, sending the observed operational data directly to the Internet of Things cloud platform. This feature enables remote observation of the gas-mixing process, preserving freshness through web browsers and mobile devices, and freeing it from geographical limitations. The system test shows that after gas mixing is completed, the gas concentration error of the mixed gases is less than 1%. The gas mixing precision is high, the operation is stable, and the process is easy to manage. This significantly reduces the cost of gas packaging equipment and extends the storage and preservation period of aquatic products, providing a reference method to improve the automation level of modified atmosphere storage and preservation for aquatic products.

The blockchain technology is sweeping the globe. Blockchain has emerged as a disruptive technology for the future generation of multiple industrial applications because to its decentralised, transparent, and secure nature. Cloud of Things, which is possible by the marriage of cloud computing with the Internet of Things, is one of them. Considering the need for security and efficiency as a problem, this paper proposes a safe and efficient smart home design that combines blockchain and cloud computing technologies to provide a comprehensive solution. The decentralised nature of blockchain technology allows it to provide processing services and create transaction copies of obtained sensible user data from smart homes. Blockchain, a distributed ledger technology that provides an immutable log of transactions recorded on a distributed network, has lately gained popularity as the underlying technology of cryptocurrencies and is revolutionising data storage and processing in computer network systems. Blockchain is seen as a possible option for future data-driven networks (DDNs) to provide safe data storage, sharing, and analytics, user privacy protection, strong, trustworthy network governance, and decentralised routing and resource management

This paper used the photoplethysmographic method to collect and extract training information data of a large number of sports personnel and preprocessed the collected data, including removing outlier and carrying out standardization processing. A non-contact network model was used to train the preprocessed data, and the physical fitness of the athletes was reflected from the following three indicators: heart rate, respiratory rate, and body temperature. This article selected 10 volunteers who love sports and compared traditional sports physiological health monitoring technology with sports physiological health monitoring technology based on the OneNet Internet of Things (IoT) cloud platform. The results show that the application of a sports physiological health monitoring system based on the IoT cloud platform can significantly improve the accuracy of physical fitness monitoring, which is of great significance for preventing safety accidents for sports personnel and improving people's body values.

Towards sustainable smart cities: Workflow scheduling in cloud of health things (CoHT) using deep reinforcement learning and moth flame optimization for edge–cloud systems

The massively generated data from various technologically advanced applications hosted in the cloud and internet of things (IoT) in present times calls for effective management towards balancing the demands of both service providers and users. The conventional usage of distributed frameworks for such big data management is witnessed with various ongoing challenges. Hence, this manuscript presents a novel analytical framework for big data that can offer reduced cost and reduced time demanded to evaluate the distributed big data from multiple data points in the cloud in an optimal way. The core ideology of this framework is to gain a synchronized optimality for cost and time for executing a task demanded for big data analytics complying with the constraints associated with task deadline. The proposed framework is capable of fine-tuning the positioning of task operation using transform and aggregate strategy to exhibit 37% reduced delay, 41% efficient task completion performance, and 28% reduced execution time in contrast to existing frameworks.

A secure and energy-efficient architecture in Internet of Things–cloud computing network by enhancing and combining three cryptographic techniques via defining new features, areas, and entities

This study designed and implemented an magnetotelluric signal acquisition and monitoring system (CMT) based on an Internet of Things (IoT) cloud platform. By integrating magnetotelluric monitoring stations, control terminals, and cloud servers, a real-time and efficient monitoring network was constructed. The hardware part of the system adopts a multi-module collaborative design, including signal conditioning circuits, FPGA control modules, DSP processing units, and embedded subsystems, achieving high-precision acquisition and processing of magnetotelluric signals. The software part employs a layered architecture, developing acquisition software, terminal control software, and a cloud platform monitoring system, which support multi-protocol communication, data parsing, and remote interaction. Through server stress testing, consistency testing, and cloud platform functional verification, the results showed that the system performs well under pressure even with limited server hardware bandwidth, with controllable consistency errors compared to the commercial device MTU-5A, and has stable field acquisition performance. The study validated the system’s advantages in real-time performance, reliability, and scalability, providing a feasible technical solution for the field of magnetotelluric monitoring. In the future, the system will be applied to geothermal monitoring.

Sports rehabilitation training is no longer simply a special sports activity. The rehabilitation ability, performance and achievements of professional athletes have not only become the hottest topics in sports competitions at all levels, but also widely regarded as the direct embodiment of the comprehensive strength of a country and a nation. Scientific monitoring of the rehabilitation training process is a powerful guarantee for scientific training and a prerequisite for good sports performance. With the development of information technology, the application fields of the Internet of Things cloud platform are more and more extensive, which can basically cover all aspects of the information technology industry. Through various information collectors, such as Radio Frequency Identification (RFID) and various sensors, it connects ubiquitous items through the network and implements information collection, transmission and processing according to agreed protocols, so as to realize the positioning, identification and tracking of ubiquitous items and finally achieve the purpose of intelligent monitoring and management of items. Starting from the practical application of sports rehabilitation training process design and monitoring, this paper establishes the system design model based on the Internet of Things cloud platform and puts forward the overall design method of the system, which makes the Internet of Things cloud platform effectively used in the sports rehabilitation training remote monitoring system.

Message digest and blockchain based chaotic ordered cyber secured cloud of things for smart health care

This paper proposes a new model that sits in the domain of improving security mechanisms in cloud and Internet of Things (IoT) utilizing deep learning based intrusion detection systems (IDS) with a sleeping stack technological, federated learning and blockchain technology. This theoretical framework seeks to address the urgent issue of protecting decentralized systems against advanced cyber threats, upholding data integrity, and maintaining privacy. Based on deep learning algorithms, the IDS detects and classifies possible security threats on a distributed network efficiently. Blockchain is used to create an immutable, transparent record of identified threats, offering solid forensic evidence and supporting decentralized, tamper-proof security protocols. In addition, federated learning is utilized in the context of our IDS models training over distributed edge nodes such that sensitive information is never shared while the models are trained, which preserves privacy per modern data protection requirements. Therefore, Preliminary experimental results show that our performance increases compared to various machine learning algorithms, as we also improve the speed of analysis, which is crucial for the intrusion detection system to react timely and minimize damage. Furthermore, the combination of blockchain and federated learning leads to improved scalability, reduced latency, and increased robustness of defense mechanisms. We demonstrate that, together the synergy of the two technologies provides a robust, scalable and privacy-respecting solution to meet the security needs of today's distributed IoT and cloud systems.

Malicious traffic segregation and attack detection caused major financial loss and became one of the most serious security hazards. Moreover, cyber security attack is the major issue, which impacts network security. The network attack methods are constantly being upgraded by the technology development and it remains a major issue for detection and protection against network attacks. For this, it is required to present an effective strategy for detecting and maintaining network security. The work provides timely and accurate congestion attack detection and identification. In the Internet of Things (IoT) cloud system malicious traffic segregation and attack detection based on a hybrid optimization-enabled deep learning (DL) network is developed in this research. At first, the input log files are gathered from the simulation of IoT sensors and the superior route is selected by the proposed Fractional Hunger Jellyfish Search Optimization (FHGJO) algorithm. The FHGJO is the integration of Hunger Game Jelly Fish Optimization (HGJO) and Fractional Calculus (FC). Furthermore, the HGJO is the combination of Hunger Game Search Optimization (HGS) with Jellyfish Optimization (JSO). Then, the segregation is done based on the fitness measures and for preprocessing; the input data is fed using quantile normalization. The feature selection process is employed using the weighted Euclidian distance (WED). With the SpinalNet, the malicious segregation is categorized as malicious and non-malicious and the proposed FHJGO is used to tune the SpinalNet. Furthermore, the proposed FHGJO-trained Deep Quantum Neural Network (DQNN) is utilized to detect the attack and classifies it into a Denial-of-Service (DOS) attack, Distributed Denial of Service (DDoS) attack, and buffer overflow attack. Moreover, the proposed model is evaluated using the NSL-KDD dataset and BoT-IoT dataset. The proposed method ensures network security with 0.931 accuracy, 0.923 sensitivity, and 0.936 specificity.

The Internet of Things (IoT) cloud model provides an efficient scheme for rapid collection, storage, processing, and analysis of massive node data, and its application has gradually expanded to key areas such as healthcare and transportation. However, the security issues of open channel transmission in IoT still persist. Researchers have proposed a lot of solutions, but the forward secrecy, session key security, and other aspects have not been effectively solved. This paper proposes a provably secure authenticated key agreement scheme, which constructs a secure channel between endpoint, gateway, and cloud server (CS). Compared with other schemes, this scheme has three characteristics: (1) According to the different computing resources of devices, gateways and CSs, a segmented differential authentication and secret key negotiation protocol is designed by using cryptographic primitives with different computing overheads; (2) after verification with the ProVerif tool, rigorous proof with the real‐or‐random (ROR) model, and informal analysis, the protocol has been proven to be secure, effectively guarding against typical threats; and (3) compared with the five most recent schemes, it can be seen that the protocol is at least 35% superior to other schemes in endpoint computational overhead, and it meets 10 security objectives, making it very suitable for application scenarios where endpoint resources are limited.

Abstract The integration of Fog Computing, Body Area Networks (BANs), and the Cloud of Things (CoT) has revolutionized smart healthcare applications, enabling real-time data processing, seamless connectivity, and efficient resource management. However, the growing demands for energy-efficient operations and reliable data transmission in these systems present significant challenges. This study proposes the development of a Deep Reinforcement Learning (DRL)-based energy-efficient routing algorithm tailored for Fog-BAN-Cloud architectures in healthcare applications. The proposed solution leverages DRL models to dynamically optimize routing paths and scheduling policies, minimizing energy consumption while maintaining high Quality of Service (QoS). The routing algorithm prioritizes low-energy paths in BAN and Fog networks. The paper specifically employs Proximal Policy Optimization (PPO), a reinforcement learning technique, to optimize the routing decisions by considering factors including energy consumption, network congestion, and data traffic conditions. PPO is used to dynamically adjust the policy updates, ensuring stability while reducing power usage and improving data transmission efficiency. Extensive simulations highlights the performance of the proposed model, demonstrating potential improvements in energy efficiency, reduced latency, and enhanced data reliability compared to traditional methods. This work highlights the potential of intelligent algorithms to address the unique challenges of healthcare-driven IoT ecosystems, providing a scalable and sustainable solution for energy-efficient routing in Fog-BAN-Cloud environments. The proposed approach is a promising strategy for optimizing IoT-driven smart healthcare systems.

This quantitative study investigated the performance, efficiency, and security resilience of quantum-resistant cryptographic protocols integrated with artificial intelligence (AI) across cloud and Internet of Things (IoT) environments. The research aimed to empirically assess whether AI-enhanced cryptographic systems could outperform conventional post-quantum algorithms in encryption throughput, latency, resource optimization, and security robustness. A factorial experimental design was implemented, encompassing multiple algorithmic classes—lattice-based, hash-based, code-based, and multivariate polynomial systems—under both AI-integrated and non-AI configurations. The analysis incorporated 7,200 experimental runs executed under varying workloads, environments, and simulated attack conditions. Linear mixed-effects models, correlation analysis, and reliability testing were used to validate the statistical integrity of the results. The descriptive analysis indicated that AI-augmented frameworks achieved consistently higher encryption speeds, lower decryption latency, and superior throughput-adjusted security efficiency compared to traditional post-quantum systems. Correlation analysis revealed strong positive relationships between AI detection accuracy, encryption performance, and system stability, confirming that AI optimization significantly improved operational consistency. Reliability and validity tests showed high internal consistency, with Cronbach’s alpha coefficients exceeding 0.90, and factor analysis confirmed that performance indicators loaded strongly on the intended theoretical constructs of cryptographic performance and AI adaptability. Collinearity diagnostics verified the independence of predictors, with all variance inflation factors below 2.0. Regression analysis demonstrated that AI integration was a statistically significant predictor of improved cryptographic outcomes (p < 0.001), increasing throughput efficiency by over 14% on average while reducing latency and energy consumption. The findings confirmed the primary hypothesis that AI-driven cryptographic optimization enhances both computational efficiency and system resilience against classical and quantum attack simulations. Lattice-based and code-based cryptosystems showed the most substantial performance gains when combined with AI learning models. Overall, the results validated that intelligent, adaptive encryption frameworks achieve measurable, statistically significant advantages in performance, scalability, and security across both cloud and IoT domains. These findings provide empirical evidence supporting the integration of AI-based decision systems into post-quantum cryptography for secure and sustainable digital infrastructures.

Chaos Game Optimization with stacked LSTM sequence to sequence autoencoder for malware detection in IoT cloud environment

Abstract The global supply chain and logistics industry is a self-organizing ecosystem consisting of numerous actors that work together to move goods from end to end. The different stakeholders involved are usually interdependent organizations, like freight forwarders, carriers, terminals, and homeland security agencies and information exchange between them is required to coordinate the activities along the individual transport chains. However, the exchange of information has often been analog, flawed, late, and incomplete. New circumstances, like unprecedented supply chain disruptions, new regulatory requirements around greenhouse gas (GHG) emissions, and generally growing shipper expectations create an urgent need for improved data sharing between actors. New technologies, like digital platforms, networks, and architectures as well as social media, mobile, analytics, cloud, and internet of things (SMACT) have brought some improvements, but not the required holistic digital perspective required or expected. Averages and approximations are usually insufficient to close data gaps, e.g., only primary data allows for accurate GHG emission calculations. Primary data sharing is widely seen as the missing piece of the puzzle. Primary data is data from the source providing an accurate state and picture of a situation. Primary data sharing at scale requires a new form of digital collaboration. We propose a rethink of digital collaboration as a means for broader primary private data sharing for complete end-to-end datasets and data quality, particularly focusing on the sharing of data associated with both transport plans and progress made in the respective movements of goods for more better disruption and carbon footprint management through more accurate calculations of estimated times of arrival (ETA) and GHG emissions. We introduce an example for digital collaboration in end-to-end supply chains that is focusing specifically on primary data sharing. The new thinking around digital collaboration manifests itself in the Virtual Watch Tower / VWT initiative (www.virtualwatchtower.org). In 2022, RISE and Singapore Maritime Institute signed a collaboration agreement focused on innovation in shipping. The VWT initiative is the first collaborative project under the umbrella of this partnership. The VWT is led by RISE, A*STAR, and VTT. VWT is a community-driven, digitally empowered initiative, a cargo owner-driven, and terminal-centric approach for improved supply chain management. It is the users themselves who co-create and co-evolve the solution that they need. The initiative aims to create a community that shapes the digital tool (VWTnet) they need to reach the required higher levels of visibility. This, through primary data sharing across the supply chain ecosystem and between actors (VWT Users) participating in individual end-to-end transports (VWT Shipments). The VWT serves as an object of research to hone the new thinking and understand the implications of its implementation.

This project will help farmers stay connected with their farm from anywhere. It can help to observe the temperature, humidity, and potential of hydrogen (pH) in the land. Most farmers will have difficulty keeping watch on the farm when traveling far away for work. The constant hot weather in Malaysia supports this. So, this project aims to design and fabricate an automatic plant watering system that helps farmers keep the farm in check at any place and time. The project will use a few sensors such as a temperature, pH, and soil humidity sensor to get the information needed and send the information to the farmers' phones through Arduino Internet of Things (IoT) Cloud. When the information has been received, farmers can activate the water sprinkler system with just a button from the phone to water the plants. The results from the project show that an automatic plant watering system has been successfully fabricated to water plants when the soil is dry. Besides, the project also successfully activates the automatic irrigation system when the soil humidity requirements are met. Finally, it is hoped that this project will help the farmer to be more efficient in keeping the farm in check and increase the quality of the product produced.