Reconfigurable intelligent surfaces (RISs) have gained considerable interest because of their inherent passive and energy-efficient design. Integrating unmanned aerial vehicles (UAVs) with reconfigurable intelligent surfaces (RIS), known as UAV-RIS, can significantly improve network performance and serve as a crucial enabler for advancements in 6G mobile networks. However, ensuring security in UAV-RIS systems poses notable challenges, particularly in the presence of imperfect channel state information (CSI) and beamforming complexities. In this paper, we identify the critical security requirements for UAV-RIS beamforming in practical scenarios. To address these challenges, we introduce a novel deep deterministic policy gradient with a distributional critic (DDPG-DC)-based beamforming approach aimed at securing UAV-RIS systems while improving the overall secrecy rate. Our proposed secure beamforming solution achieves up to a 48% performance improvement compared to existing state-of-the-art algorithms.

Unmanned aerial vehicles (UAVs) in integration with reconfigurable intelligent surfaces (RIS) play a crucial role in improving wireless communication coverage and enhancing overall performance. However, optimizing the beamforming for the RIS and base station (BS) is critical in improving coverage and ensuring connectivity in densely populated areas. The primary challenge in this process arises from the diverse Quality of Service (QoS) requirements set by user equipment (UEs). To address this complexity, machine learning algorithms are employed to predict the optimal beamforming configurations for both the BS and RIS. However, the traditional supervised learning methods are becoming less effective due to the ever-changing demands of UEs, as these methods rely on fixed data patterns that struggle to adapt to the fluctuating QoS requirements of UEs. Thus, in this paper, we propose an unsupervised learning-based deep learning (DL) approach to jointly predict the optimal beamforming matrix for RIS and BS, enhancing communication coverage and maximizing QoS satisfaction of UEs. The proposed DL-based beamforming adaptively predicts the beamforming matrix, facilitates efficient data exploration during the initial learning phase, and seamlessly scales as the process advances, thereby enhancing overall performance. Numerical results demonstrate that the proposed DL-based RIS and BS beamforming outperforms by up to 89%, compared to the state-of-the-art methods.

Abstract Numerous techniques have been proposed in the literature to improve the performance of high‐gain observers with noisy measurements. One such technique is the linear extended state observer, which is used to estimate the system's states and to account for the impact of internal uncertainties, undesirable nonlinearities, and external disturbances. This observer's primary purpose is to eliminate these disturbances from the input channel in real‐time. This enables the observer to precisely track the system states while compensating for the various sources of uncertainty that can influence the system's behaviour. So, in this paper, a novel nonlinear higher‐order extended state observer (NHOESO) is introduced to enhance the performance of high‐gain observers under noisy measurement conditions. The NHOESO is designed to observe the system states and total disturbance while eliminating the latter in real time from the input channel. It is capable of handling disturbances of higher‐order derivatives, including internal uncertainties, undesirable nonlinearities, and external disturbances. The paper also presents two innovative schemes for parametrizing the NHOESO parameters in the presence of measurement noise. These schemes are named time‐varying bandwidth NHOESO (TVB‐NHOESO) and online adaptive rule update NHOESO (OARU‐NHOESO). Numerical simulations are conducted to validate the effectiveness of the proposed schemes, using a nonlinear uncertain system as a test case. The results demonstrate that the OARU technique outperforms the TVB technique in terms of its ability to sense the presence of noise components in the output and respond accordingly. However, it is noted that the OARU technique is slower than the TVB technique and requires more complex parameter tuning to adaptively account for the measurement noise.

ABSTRACT This study introduces an innovative optimization algorithm called Lionfish Search (LFS) technique, which is inspired by the visual predator Lionfish, in which it is specifically imitating their hunting tactics. The suggested algorithm considers several parameters that influence the hunting behaviour of lionfish, such as visual acuity, mobility, striking success, and prey swallowing potential. Furthermore, this study examines the influence of the physiological traits of the lionfish and their relationship with environmental factors. The novel search algorithm has shown enhanced performance and efficiency, particularly in scenarios where the integration of visual cues and intricate hunting strategies is vital. The suggested LFS method was evaluated using 20 well‐known single‐modal and multi‐modal mathematical functions to analyse its different characteristics. The LFS method has shown remarkable efficacy in both exploration and exploitation, effectively reducing the likelihood of being trapped in local optima. Additionally, it has a rapid convergence capacity, particularly in the realm of large‐scale global optimization. Comparisons were made between the LFS algorithm, and 10 other prominent algorithms mentioned in the literature. The proposed LFS metaheuristic algorithm outperformed the others on almost all of the examined functions, demonstrating a statistically significant advantage. Moreover, the positive results found in three practical optimization situations demonstrate the effectiveness of the LFS in accomplishing problem‐solving tasks that have limited and unknown search areas.

Abstract From a future perspective and with the current advancements in technology, deep reinforcement learning (DRL) is set to play an important role in several areas like transportation, automation, finance, medical and in many more fields with less human interaction. With the popularity of its fast‐learning algorithms there is an exponential increase in the opportunities for handling dynamic environments without any explicit programming. Additionally, DRL sophisticatedly handles real‐world complex problems in different environments. It has grasped great attention in the areas of natural language processing (NLP), speech recognition, computer vision and image classification which has led to a drastic increase in solving complex problems like planning, decision‐making and perception. This survey provides a comprehensive analysis of DRL and different types of neural network, DRL architectures, and their real‐world applications. Recent and upcoming trends in the field of artificial intelligence (AI) and its categories have been emphasized and potential challenges have been discussed.

Abstract The vulnerabilities of the Internet of Things (IoTs) in general and the Internet of Mobile Things (IoMTs) in particular motivate researchers to equip them with security systems against intruders and attacks. The integration of anomaly detection with intrusion detection for IoMTs has not been addressed adequately. This paper tackles this issue through building a Kalman filter and Cauchy clustering algorithm for anomaly detection and using them for authentication nodes within IoMTs using the Extreme Learning Machine classifier. The algorithm of this proposed work is composed of various components; first, the Kalman filter‐based model for estimating the trajectory of pedestrians within an indoor environment based on fusing WiFi with IMU data. Second, trustworthiness assessment for detecting anomaly behaviour in IoMT based on the estimated trajectory using the Kalman filter. Third, the trust IDS model for IoMT systems by integrating anomaly detection with online learning for attacks identification using an online sequential extreme learning machine. The algorithm has been implemented and evaluated using TamperU dataset for WiFi fingerprinting and KDD99 for intrusion detection. Furthermore, a comparison with benchmarks (the algorithms which used in other studies) for intrusion and anomaly detection proves the superiority of this proposed approach in terms of all the considered classification metrics.

From a Licensed Medical Practitioner’s (LMP) perspective, e-Healthcare Risk Prediction plays a vital role in Health Big Data. This also is a hot issue in e-healthcare because of the lack of security and privacy protections. To overcome this deficiency, this research article proposes heterogeneous network systems (HNS), an efficient and privacy-preserving e-Healthcare Risk Prediction method for e-healthcare. In comparison to the existing research contribution, the proposed HNS accomplish two steps of disease risk prediction, namely Analysis of HNS, and Heterogeneous Network (HetNet) concerning the LMP for analyzing the in-hospital involvement care by collecting and explaining the “Health Big Data” as per the view of the LMP. This will help to access the services from the hospital. In the LMP-Centric Heterogeneous Network Powered Efficient e-Healthcare Risk Prediction phase, the “Polygenic Score” is calculated for risk prediction for health big data. Through the characteristics of “non-predictive applications” and “Predictive applications,” procedural aspects are analyzed with the LMP-Centric HetNet against the Efficient e-Healthcare Risk Prediction. This will be applied to the Medical extensive data integration and clustering for handling Health Big Data. Finally, the LMP-Centric HetNet Powered Efficient e-Healthcare Risk Prediction for Health Big Data treats the LMP perspective efficiently. The proposed system increased prediction accuracy to 45.9%, and the monogenic score increased from 3% to 19%. The density accuracy range is increased from 13.9% to 39%. The increased execution time is improved from 29.95% to 36.05%. This comprehensive prediction analysis accuracy range is 73.98% efficient.

Abstract Dealing with imbalanced data is a common challenge in machine learning, where one class has significantly fewer examples than another. Successfully addressing this challenge requires careful consideration of the data, algorithm, and evaluation metrics to ensure that the model accurately predicts the minority class. In this study, we present a hybrid approach called Siamese‐HYNAA, which combines a Siamese network and a population‐based optimizer hypercube natural aggregation algorithm (HYNAA) to generate candidate solutions for augmenting the minority class. We collected 10 imbalanced datasets ranging from 1.81 to 8.78 imbalanced ratios and built solution pairs based on correctly predicted candidate solutions using support vector machine (SVM). We then fed these solutions to the Siamese network, which employs a one‐shot learning approach to improve predictions with fewer candidate solutions. However, we found that SVM predicted only a small number of minority class samples accurately, prompting us to optimize the number of candidate solution pairs using HYNAA to generate more synthetic samples for the Siamese network. We evaluated our proposed strategy against basic SMOTE and our previous work, SMOTE‐PSOEV, using various performance measures, including ROC‐AUC learning curves, sensitivity, specificity, accuracy, Characteristic stability index, balanced accuracy, F1‐score, informedness, markedness, and execution time. Our results indicate that Siamese‐HYNAA generates promising results for imbalanced data.

Nowadays, Green IoT-Based Agriculture plays an essential role in farming to improve the yield. Here, IoT devices are embedded in the farming equipment, which helps to enhance the irrigation and yield with minimum cost-cutting. Data security and privacy are major challenges in green IoT-related agriculture. Therefore, a secured system should create to maintain data confidentiality, authentication, integrity, availability, and privacy. This system uses the privacy-preserving data aggregation (PPDA) with a Fair access framework (FAF) that manages the data security. The data aggregation concept is used to protect the green IoT data from false data injection. The FAF utilizes the blockchain technique to grant, get, revoke and delegate access to the user. The developed security system can adapt the green IoT-based agriculture and provide confidentiality, which is done with the help of an enhanced ciphertext access control mechanism. This system resolves the security and privacy issues involved in the Green IoT-based agriculture, and the effectiveness of the system is evaluated using implementation results.

Head movement holds significant importance in con-veying body language, expressing specific gestures, and reflecting emotional and character aspects. The detection of head movement in smart or assistive driving applications can play an important role in preventing major accidents and potentially saving lives. Additionally, it aids in identifying driver fatigue, a significant contributor to deadly road accidents worldwide. However, most existing head movement detection systems rely on cameras, which raise privacy concerns, face challenges with lighting conditions, and require complex training with long video sequences. This novel privacy-preserving system utilizes UWB-radar technology and leverages Deep Learning (DL) techniques to address the mentioned issues. The system focuses on classifying the five most common head gestures: Head 45L (HL45), Head 45R (HR45), Head 90L (HL90), Head 90R (HR90), and Head Down (HD). By processing the recorded data as spectrograms and leveraging the advanced DL model VGG16, the proposed system accurately detects these head gestures, achieving a maximum classification accuracy of 84.00% across all classes. This study presents a proof of concept for an effective and privacy-conscious approach to head position classification.

Abstract Because of increased applications of Internet of Things (IoT) environment in real‐time environment, confidential data gathered by the IoT devices are being communicated to the cloud environment to train the machine learning (ML) models in understanding the patterns that exist in the data. At the same time, the sensitive nature of the IoT data attracts malicious users into hacking efforts. An intrusion detection system (IDS) can be applied to ensure security in the IoT environment. In order to improve security, the ML models can be executed at the data source instead of centralized cloud server. Federated learning (FL) is a recent progression of ML model which enables the ML models to move into the data source rather than moving the data to centralized cloud and thereby resolves cybersecurity problems in the IoT environment. In this view, this study introduces an FL based IDS using bird swarm algorithm based feature selection with classification (FLIDS‐BSAFSC) model in IoT environment. The presented FLIDS‐BSAFSC model undergoes training on multiple aspects of IoT dataset in a decentralized format to classify, detect, and defend against attacks. The proposed FLIDS‐BSAFSC model initially applies min–max normalization technique to pre‐process the IoT data. Besides, BSA based feature selection (BSA‐FS) technique is designed to elect feature subsets. Finally, social group optimization algorithm with kernel extreme learning machine model is employed for identifying various kinds of classes. In the view of FL where the IoT dataset is not distributed to the server carries out profile aggregation competently with the advantage of peer learning. The experimental validation of the FLIDS‐BSAFSC model is tested using benchmark datasets and the results are inspected under several aspects. The experimental values highlighted the better performance of the FLIDS‐BSAFSC model over recent approaches.

Electronic health records are essential and sensitive since they include vital information and are routinely exchanged across several parties, such as hospitals and private clinics. These data must remain accurate, current, secret, and available only to authorized parties. Integrating these data improves the accuracy and cost-effectiveness of the present health data administration framework. Electronic Medical Records (EMRs) are now kept utilizing the structure of the client/server via whom patient data information is maintained in the hospital. Multiple hospitals use the same database to track a single patient. These limitations prevent a custom health system from providing various associated experts and patients with a cohesive, integrated, secure, and confidential medical history. Modern healthcare systems are distinguished by their complexity and expense. However, this may be mitigated by enhanced health record management and Blockchain technology. The Blockchain’s data availability, confidence, and security characteristics have a bright future in healthcare services, giving solutions to the issues of the traditional customer/server architecture EMR management platform: intricacy, confidence, dependability, compatibility, and anonymity. An e-health record management based on Internet of Medical Things (EHRM-IoMT) is proposed in this paper. This paper explores and analyzes Blockchain efficiency and customer/server paradigms. The findings show that a patient-centred strategy may achieve remarkable success utilizing Blockchain. Moreover, the immutable and accurate data of persons in Blockchain may enable healthcare practitioners to better forecast and aid with diagnosis utilizing the IoMT via machine learning and artificial intelligence.

Road traffic crashes are one of the prime issues in the world. Every year 1.35 million people die, 20–50 million fatal injuries, and many incur a disability due to road traffic crashes. 50% of the total death, injuries, and disabilities are among vulnerable road users (VRUs), such as motorcyclists, cyclists, and pedestrians. Among these VRUs, highly vulnerable is pedestrians. In this article, we are dealing with providing safety and alert system to pedestrians and vehicles to reduce and/or avoid the causes of road traffic crashes in metropolitan areas. In this article, we develop a cooperative communication framework between pedestrians and nearby vehicles as well as traffic light systems using mobile agent systems and apps. The proposed cooperative communication framework controls the mobility of vehicles as well as pedestrians to avoid accidents. We have developed an application that will notify the pedestrian if there is any automobile within the range of 200 m of the pedestrian. If there are any vehicles in this range, using direct Wi-Fi, connect the vehicles and pedestrians to notify them about their presence of them. The proposed system is implemented and tested in real time as well as simulated in the SUMO, Veins, OMNeT++, and MiXiM simulator. The proposed system’s results (real time, simulation, and comparison) show real-time deployability, accuracy, and reliability.

These days, the usage of machine-learning-enabled dynamic Internet of Medical Things (IoMT) systems with multiple technologies for digital healthcare applications has been growing progressively in practice. Machine learning plays a vital role in the IoMT system to balance the load between delay and energy. However, the traditional learning models fraud on the data in the distributed IoMT system for healthcare applications are still a critical research problem in practice. The study devises a federated learning-based blockchain-enabled task scheduling (FL-BETS) framework with different dynamic heuristics. The study considers the different healthcare applications that have both hard constraint (e.g., deadline) and resource energy consumption (e.g., soft constraint) during execution on the distributed fog and cloud nodes. The goal of FL-BETS is to identify and ensure the privacy preservation and fraud of data at various levels, such as local fog nodes and remote clouds, with minimum energy consumption and delay, and to satisfy the deadlines of healthcare workloads. The study introduces the mathematical model. In the performance evaluation, FL-BETS outperforms all existing machine learning and blockchain mechanisms in fraud analysis, data validation, energy and delay constraints for healthcare applications.

Wireless technology and sensing devices are playing an important role in healthcare, known as Body Sensor Networks (BSNs). Existing wearable technologies process vast amounts of data with critical quality of service (QoS) requirements in terms of delay, reliability, and throughput. This study provides a traffic prioritizing strategy that ensures synchronization, optimal traffic control, and resource optimization. It includes a method for reducing delay and enhancing throughput, and the energy efficiency of BSNs. In addition, we investigated that implementation of access periods improves the channel accessing strategy for high priority nodes with increased starvation for high data rates in low priority nodes. M/G/1/K queue with finite buffer is implemented to overcome poor resource utilization. Simulation results showed that implementing a finite buffer had enhanced resource utilization in terms of higher throughput and bandwidth efficiency.

Abstract Driven by the requirements for entirely low communication latencies, high bandwidths, reliability and capacities, the Fifth Generation (5G) networks has been deployed in a number of countries. One of the most prevalent application scenarios of 5G networks is the Internet of Things (IoT) that can potentially boost convenience and energy savings. However, the information exchanged over the open wireless 5G networks is susceptible to numerous attacks such as malicious modifications. Although many protocols have been developed to protect against these attacks, the provision of optimum security and privacy issues in 5G networks is still an open challenge. This is attributed to the high device density, frequent handovers and resource constrained nature the 5G IoT nodes. In this article, a network selection and authentication protocol that securely verifies the authenticity of all the communicating entities is presented. The network selection is accomplished using Artificial Neural Network (ANN) for increased efficiency. In addition, all the security tokens are independently derived at the end devices without the involvement of any central authority. Formal security analysis based on the Burrows–Abadi–Needham (BAN) logic shows that all the terminals securely authenticate each other before the onset of packet exchanges. In addition, it is shown that this protocol thwarts majority of the conventional 5G attack vectors and is robust under the Dolev–Yao (DY) threat model. Moreover, a comparison with other related schemes shows that the proposed protocol offers many adorable security features at relatively low communication and computation costs. The simulation results show that the deployed ANN yields low packet loss ratio and latency variations.

The advancement of healthcare technology is impossible without machine learning (ML). There have been numerous advances in ML to analyze, predict, and diagnose medical data. Integrating a centralized scheme and therapy for classifying and diagnosing illnesses and disorders is a major obstacle in modern healthcare. To standardize all medical data into a single repository, researchers have proposed using ML using the centralized artificial neural network model (ML-CANNM). Random tree, support vector machine, and gradient booster are just a few proposed ML classifiers. Artificial neural networks (ANNs) have been trained using a variety of medical datasets to predict and analyze outcomes. ML-CANNM collects patient data from various studies and uses ML and ANNs to determine the results. Three layers make up an ANN. ML is used to classify the given patients’ data in the input layer. In the hidden layer, classification data are compared to a training dataset. The output layer’s job is to identify, classify, and diagnose diseases. As a result, disease diagnosis and detection are integrated into a single healthcare database. The proposed framework has proven that ML-CANNM works with more accuracy and lesser execution time. Thus, the numerical outcome suggested ML-CANNM increased accuracy ratio of 99.2% and a prediction ratio of 97.5%. The findings further show that the execution time is enhanced by less than 2[Formula: see text]h, decision table using ML and results in an efficiency ratio of 97.5%.

With the widespread utilization of Internet-enabled smart devices (SDs), the Industrial Internet of Things (IIoT) has become prevalent in recent years. SDs exchange information through the open Internet, which creates security and privacy concerns for the exchanged information. To address these concerns, various solutions exist in the literature which, because of high computational and communication overheads, are not appropriate for the resource-constricted IIoT environment. This article proposes a resource-efficient authentication protocol for the IIoT, called REAP-IIoT, which employs a lightweight cryptography (LWC)-based authenticated encryption with associative data (AEAD) primitive AEGIS along with hash function. LWC-based AEAD primitives are suitable for resource constraint SDs because they require fewer computational resources. Moreover, REAP-IIoT renders the privacy-preserving user authentication functionality and establishes a session key (SK) between SDs deployed in the IIoT environment and users. Both user and SD utilize the established SK for encrypted communication. The security of SK, established during the authentication and key exchange (AKE) process of REAP-IIoT, is validated through the broadly accepted random or real model. Besides, Scyther-based security verification is conducted to illustrate that REAP-IIoT is secure and can protect the man-in-the-middle and replay attacks. Additionally, the informal security analysis is carried out to show that REAP-IIoT is protected against various covert security risks. A thorough comparison reveals that REAP-IIoT renders enhanced security characteristics apart from its low communication, storage, and computational overheads than the relevant AKE protocols.

The biological human brain model was used to inspire the idea of Artificial Neural Networks (ANNs). The notion is then converted into a mathematical formulation and then into machine learning, which is utilized to address various issues throughout the world. Moreover, ANNs has achieved advances in solving numerous intractable problems in several fields in recent times. However, its success depends on the hyper-parameters it selects, and manually fine-tuning them is a time-consuming task. Therefore, automation of the design or topology of artificial neural networks has become a hot issue in both academic and industrial studies. Among the numerous optimization approaches, evolutionary algorithms (EAs) are commonly used to optimize the architecture and parameters of ANNs. We review several successful, well-designed strategies to using EAs to develop artificial neural network architecture that has been published in the last four years in this paper. In addition, we conducted a thorough study and analysis of each publication. Furthermore, details such as methods used, datasets, computer resources, training duration, and performance are summarized for each study. Despite this, the automated neural network techniques performed admirably. However, the long training period and huge computer resources remain issues for these sorts of ANNs techniques.