Federated Learning for IoUT: Concepts, Applications, Challenges and Future Directions

The development of intelligent systems for the monitoring, exploration, and administration of underwater environments is made possible by the Internet of Underwater Things (IoUT), which is a revolutionary development in marine and environmental research. This thorough research examines the advancements, challenges, and promise of IoUT with a focus on its applications in domains such as resource extraction, the science of oceanography underwater tracking, and tracking the environment. IoUT systems require customised approaches in processing information, conservation of energy, connectivity, and sensor that is being tested design because to the particular difficulties of underwater settings. The new protocols and methods designed for underwater applications—such as acoustic, optical, and electromagnetic communications—as well as the incorporation of artificial intelligence (AI) and machine learning (ML) technologies for improved data processing and taking decisions are covered in this study. Furthermore, we draw attention to the urgent issues surrounding data security, environmentally friendly interaction, and installation expenses while providing suggestions for future lines of inquiry and technical advancements.

Within the domain of password security classification, the pursuit of practical and dependable methodologies has prompted the examination of both biological and technological paradigms. The present study investigates the efficacy of Mouth Brooding Fish (MBF) as an innovative method in contrast to conventional Machine Learning (ML) approaches for classifying password security. The research approach entails a rigorous examination of the comparative analysis of MBF and ML algorithms, evaluating their effectiveness in password classification using many criteria, including accuracy, robustness, flexibility, and durability against adversarial assaults. The findings suggest that ML approaches have shown significant effectiveness in classifying passwords. However, using methodologies inspired by the minimum Bayes risk framework demonstrates a higher degree of resistance against typical cyber dangers. The intrinsic biological mechanisms of MBF, encompassing adaptive behaviors and inherent protection, play a role in enhancing the resilience and adaptability of the password security categorization system. The results offer significant insights that can inform the evolution of password security systems, integrating biological principles with technical progress to enhance safeguarding measures in digital environments. To emphasize the advantages of the suggested approach, several ML approaches are investigated, such as Support Vector Machines (SVM), AdaBoost, Multilayer Perceptron (MLP), Gaussian Kernel (GK), and Random Forest (RF). The F-score, accuracy, sensitivity, and specificity metrics for MBF exhibit noteworthy performance compared to the other selected models, with values of 100%.

In recent years, along with the blooming of Machine Learning (ML)-based applications and services, ensuring data privacy and security have become a critical obligation. ML-based service providers not only confront with difficulties in collecting and managing data across heterogeneous sources but also challenges of complying with rigorous data protection regulations such as EU/UK General Data Protection Regulation (GDPR). Furthermore, conventional centralised ML approaches have always come with long-standing privacy risks to personal data leakage, misuse, and abuse. Federated learning (FL) has emerged as a prospective solution that facilitates distributed collaborative learning without disclosing original training data. Unfortunately, retaining data and computation on-device as in FL are not sufficient for privacy-guarantee because model parameters exchanged among participants conceal sensitive information that can be exploited in privacy attacks. Consequently, FL-based systems are not naturally compliant with the GDPR. This article is dedicated to surveying of state-of-the-art privacy-preservation techniques in FL in relations with GDPR requirements. Furthermore, insights into the existing challenges are examined along with the prospective approaches following the GDPR regulatory guidelines that FL-based systems shall implement to fully comply with the GDPR.

Performance optimization of wireless networks is typically complicated because of high computational complexity and dynamic channel conditions. Considering a specific case, the recent introduction of intelligent reflecting surface (IRS) can reshape the wireless channels by controlling the scattering elements' phase shifts, namely, passive beamforming. However, due to the large size of scattering elements, the IRS's beamforming optimization becomes intractable. In this article, we focus on machine learning (ML) approaches for complex optimization problems in wireless networks. ML approaches can provide flexibility and robustness against uncertain and dynamic systems. However, practical challenges still remain due to slow convergence in offline training or online learning. This motivated us to design a novel optimization-driven ML framework that exploits the efficiency of model-based optimization and the robustness of model-free ML approaches. Splitting the control variables into two parts allows one part to be updated by the outer loop ML approach while the other part is solved by the inner loop optimization. The case study in IRS-assisted wireless networks confirms that the optimization-driven ML framework can improve learning efficiency and the reward performance significantly compared to conventional model-free ML approaches.

CLAIM is a tool for analyzing and classifying discrepancy reports that allows for the incorporation of domain expert knowledge into a semi-supervised machine learning (ML) process (a semi-supervised learning uses a small number of manually labeled data and a much larger amount of unlabeled for training a machine learning algorithm). By using this domain knowledge, classification accuracy is higher than conventional ML approaches. The advantages are particularly apparent with small, imbalanced data sets that are quite common in discrepancy report data sets (an imbalanced data set has an unequal distribution of documents categories within each category). The CLAIM method is robust against human bias and can tolerate misclassifications of up to 20% of the training set. The increased accuracy of the CLAIM methodology makes ML a viable tool for safety, reliability, and software development process decision making. The modest human labor requirement enables use of the method under circumstances that previously made free text discrepancy report analysis infeasible due to resource, scheduling, and cost constraints.

The Internet of underwater things (IoUT) has become a promising and vibrant paradigm which mainly relies on the underwater acoustic wireless sensor networks (UASN). IoUT was first proposed to facilitate offshore real‐time monitoring and exploration of underwater environments. It is now becoming a key component for building and connecting the futuristic underwater smart cities. However, Acoustic waves used in underwater communication suffer from long propagation delays and high transmission energy requirements. Therefore, approaches for autonomous underwater vehicle (AUV) assisted data collection and cluster‐based routing were proposed to overcome these drawbacks. In this article, we present a smart genetic‐based AUV path planning algorithm for data collection to enhance the performance of UASN. In addition, we combine our smart data collection technique with a dynamic location‐unaware clustering algorithm to further reduce energy consumption with mobility consideration. Our simulation results showed that our genetic‐based technique achieves more than 145% increase in network lifetime and almost a 75% reduction in energy expenditure compared with traditional diagonal path selection technique. More simulations comparing our combined technique with a novel clustering technique showed that the network lifetime is twice as better, total energy consumption is nearly 30% less, and the delivery ratio is 15% more at low densities.

Data-driven or machine learning (ML) approaches already achieved significant success in many engineering areas even fatigue assessment of industrial parts and structures. Machine learning approaches work well under the common assumption that the training data covers the relevant feature space of the application. Rebuilding new models or establish new databases for similar feature spaces needs a high effort. In such cases, knowledge transfer or transfer learning can be used. In this study, transfer learning approach is used to determine the fatigue life of welded steel joints (target task) with a ML-algorithm that is trained in non-welded steel specimen. Twenty-two fatigue test data series were used. The results of the transfer learning approach were compared with a conventional machine learning approach that was trained also on data of welded joints. Furthermore, the results were compared to an advanced analytical approach (IBESS) for the fracture mechanic-based fatigue life assessment of welded joints and fatigue strength values from recommendations.

An emerging technology with the capacity to revolutionize the transportation sector is autonomous driving, offering the promise of heightened safety, efficiency, and convenience. However, the widescale deployment of autonomous vehicles presents a multitude of challenges, notably the necessity for robust and adaptable machine learning (ML) models capable of handling a wide array of dynamic real-world scenarios. Enter Federated Learning (FL), a decentralized ML approach that has gained recognition as a potential solution to these challenges. This paper delves into the primary advantages of FL within the context of autonomous driving. It highlights FL's capacity to seamlessly adapt to edge devices, respond to localized changes, and continually enhance safety and performance. The document substantiates these advantages through numerous case studies and empirical evidence, demonstrating how FL can potentially elevate the vision, decision-making, control systems, data transmission, and learning model capabilities of autonomous vehicles. By harnessing the collective intelligence of autonomous vehicles while preserving data privacy and security, FL holds the potential to propel us closer to a future where safe, efficient, and autonomous transportation becomes an attainable reality.

Obstacle detection on the railway, a crucial operational safety concern, is a complex task that encompasses a multitude of challenges. While Machine Learning (ML) algorithms are commonly employed in analogous applications such as autonomous car driving [1] [2], the railway field faces a significant barrier due to the scarcity of available data (particularly images), rendering conventional ML approaches impractical. In response to this challenge, this study proposes and evaluates a framework which uses LiDAR (Light Detection and Ranging) data for obstacle detection on the railways. The framework aims to address the limitations posed by image data scarcity while enhancing operational safety in railway environments. The developed methodology combines the use of a long-range LiDAR capable of detecting obstacles at distances of up to 500 meters, with the train’s GPS (Global Positioning System) coordinates to accurately determine its position relative to detected obstacles. The LiDAR data is processed using a data fusion approach, where pre-existing knowledge regarding the track topography is combined with a clustering algorithm, specifically DBSCAN (Density-based spatial clustering of applications with noise), to identify and classify potential obstacles at a pre-defined distance. Tests of the proposed framework were conducted within the confines of a moving locomotive, specifically the CP 2600-2620 series, along a designated section of the Contumil-Leixões line. These tests served to validate the effectiveness and feasibility of the approach under real-world operating conditions. Overall, the utilization of LiDAR data coupled with advanced algorithms presents a promising avenue for enhancing obstacle detection capabilities in railway operations. By overcoming the challenges associated with data scarcity, this framework holds the potential to significantly improve operational safety and efficiency within railway networks. Further research and testing are warranted to validate the framework’s performance across diverse railway environments and operating conditions.

The healthcare sector is increasingly leveraging Internet of Things (IoT) technologies to enable real-time patient monitoring, predictive diagnostics, and personalized treatment. However, the large volume of heterogeneous data generated by IoT devices poses challenges in data processing, accuracy, and timely decision-making. This study proposes an IoT optimization system using machine learning (ML) to enhance the efficiency and reliability of healthcare monitoring systems. The framework integrates wearable sensors, smart medical devices, and cloud/edge computing platforms to collect and preprocess physiological and environmental data from patients. These devices generate large volumes of sensitive patient data that require secure, efficient, and privacy-preserving analytics. Conventional cloud-based machine learning approaches often expose raw data to centralized servers, creating significant challenges in terms of data privacy, network congestion, latency, and compliance with healthcare regulations. To address these challenges, this thesis proposes an IoT Device Optimization System using Federated Learning (FL) for intelligent, distributed data processing and device performance enhancement in a healthcare environment. A hybrid Agile–Incremental development methodology was adopted to support iterative prototyping, modular expansion, and continuous validation. Comprehensive testing including unit, integration, system, and performance evaluation demonstrated that the proposed system improves prediction accuracy, minimizes data exposure risk, and enhances device responsiveness. The results show that FL-enabled optimization improves model accuracy (93%), and maintain strict privacy standards. This thesis contributes a scalable, secure, and intelligent framework for optimizing healthcare IoT devices, supporting remote diagnostics, smart clinical environments, and personalized healthcare delivery.

The Internet of Underwater Things (IoUT) is an evolving class of Internet of Things and it is considered the basic unit for the development of smart cities. To support the idea of IoUT, an Underwater Sensor Network (USN) has emerged as a potential technology that has attractive and updated applications for underwater environment monitoring. In such networks, route selection and cluster-head management are still challenging. As the optimal routes always lead to congestion and longer delays while the cluster-head mismanagement leads to ending the USN lifespan earlier. In this paper, we propose a cooperative clustering game that is based upon energy heterogeneity and a penalty mechanism to deal with the cluster head mismanagement issue. Then, we use a non-cooperative evolutionary game for the best relay selection; the results prove that this utility function is the most suitable solution for the relay selection and its strategy selection converges to Nash Equilibrium. The proposed framework is compared with recent schemes using different quality measures and we found that our proposed framework performs favorably against the existing schemes for all of the evaluation metrics.

This systematic literature review uses the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to assess progress in blockchain-based Federated learning (FL) and Machine Learning (ML) for detecting financial fraud over the last five years (2020–2025). An initial pool of 29,274 records identified across IEEE Xplore, ACM Digital Library, and ScienceDirect yielded 1585 peer-reviewed studies that met the inclusion criteria. Both qualitative and quantitative approaches were used. The examined papers were classified according to algorithm type, fraud types, and evaluation measures. Credit card fraud and cryptocurrency fraud dominated the literature, with supervised learning (e.g., XGBoost, 95% accuracy) and federated learning (e.g., FedAvg, 91% accuracy) emerging as dominant methodologies. Centralized ML outperforms FL in latency but poses privacy risks. FL–blockchain hybrids reduce false positives. While precision, recall, and F1-score are commonly used, few studies use cost-sensitive criteria. Future research should prioritize adaptive FL aggregation, privacy-preserving ML, and cross-industry collaboration.

Federated Learning (FL) is a new approach in machine learning or it can also be called collaborative learning, which is a machine learning method that includes client devices to carry out the training process, so that clients do not need to send training data to the server but directly conduct training on their respective devices. respectively. The models generated from local training will be sent to the server for further global aggregation. Therefore, FL is referred to as machine learning which can maintain the privacy of the data owner, because the data is not submitted to the server and is still stored in each client's device. In this study, a performance comparison will be carried out to prove whether the latest approach which is Federated Learning, can produce the same accuracy performance as the traditional approach, that is Centralized Machine Learning, in the case of Image Classification. The comparison of two approaches would be conducted by using the Open Source Image Classification dataset, namely MNIST. The performance of two approaches would be presented by evaluation that is Accuracy. The result shows that Federated Learning almost overcome the performance of Centralized Learning in the case of Image Classification by provided Accuracy 76%.

Underwater is a harsh and dynamic environment. Its physical properties are unsteady in the space-time domain and that complicates the communication process inside the water. Moreover, underwater transmissions suffer from many attenuations and scattering sources due to several environmental noises such as shipping, turbulence, wave dynamics, and the background life of the marine organisms.In this chapter, we introduce the underwater communication basics and challenges compared with terrestrial networks, then illustrate the Internet of Underwater Things (IoUT) network structure and the underwater node architecture, and finally, overview the machine learning modeling in underwater communication.KeywordsUnderwaterInternet of Underwater Things (IoUT)Cooperative communicationMachine learning modelingUnderwater wireless sensor networksAmplify-and-forward (AF)Decode-and-forward (DF)Hybrid and cooperative transmissionTransmission power levelsRelayMoteMachine learning techniquesSupervised learningMachine learning algorithms

Chemical heuristics have been fundamental to the advancement of chemistry and materials science. These heuristics are typically established by scientists using knowledge and creativity to extract patterns from limited datasets. Machine learning offers opportunities to perfect this approach using computers and larger datasets. Here, we discuss the relationships between traditional heuristics and machine learning approaches. We show how traditional rules can be challenged by large-scale statistical assessment and how traditional concepts commonly used as features are feeding the machine learning techniques. We stress the waste involved in relearning chemical rules and the challenges in terms of data size requirements for purely data-driven approaches. Our view is that heuristic and machine learning approaches are at their best when they work together.