AI and Blockchain-Based Secure Data Dissemination Architecture for IoT-Enabled Critical Infrastructure.

The Internet of Things (IoT) is the most abundant technology in the fields of manufacturing, automation, transportation, robotics, and agriculture, utilizing the IoT's sensors-sensing capability. It plays a vital role in digital transformation and smart revolutions in critical infrastructure environments. However, handling heterogeneous data from different IoT devices is challenging from the perspective of security and privacy issues. The attacker targets the sensor communication between two IoT devices to jeopardize the regular operations of IoT-based critical infrastructure. In this paper, we propose an artificial intelligence (AI) and blockchain-driven secure data dissemination architecture to deal with critical infrastructure security and privacy issues. First, we reduced dimensionality using principal component analysis (PCA) and explainable AI (XAI) approaches. Furthermore, we applied different AI classifiers such as random forest (RF), decision tree (DT), support vector machine (SVM), perceptron, and Gaussian Naive Bayes (GaussianNB) that classify the data, i.e., malicious or non-malicious. Furthermore, we employ an interplanetary file system (IPFS)-driven blockchain network that offers security to the non-malicious data. In addition, to strengthen the security of AI classifiers, we analyze data poisoning attacks on the dataset that manipulate sensitive data and mislead the classifier, resulting in inaccurate results from the classifiers. To overcome this issue, we provide an anomaly detection approach that identifies malicious instances and removes the poisoned data from the dataset. The proposed architecture is evaluated using performance evaluation metrics such as accuracy, precision, recall, F1 score, and receiver operating characteristic curve (ROC curve). The findings show that the RF classifier transcends other AI classifiers in terms of accuracy, i.e., 98.46%.

Brain Tumor Classification from MRI Using Image Enhancement and Convolutional Neural Network Techniques.

The independent detection and classification of brain malignancies using magnetic resonance imaging (MRI) can present challenges and the potential for error due to the intricate nature and time-consuming process involved. The complexity of the brain tumor identification process primarily stems from the need for a comprehensive evaluation spanning multiple modules. The advancement of deep learning (DL) has facilitated the emergence of automated medical image processing and diagnostics solutions, thereby offering a potential resolution to this issue. Convolutional neural networks (CNNs) represent a prominent methodology in visual learning and image categorization. The present study introduces a novel methodology integrating image enhancement techniques, specifically, Gaussian-blur-based sharpening and Adaptive Histogram Equalization using CLAHE, with the proposed model. This approach aims to effectively classify different categories of brain tumors, including glioma, meningioma, and pituitary tumor, as well as cases without tumors. The algorithm underwent comprehensive testing using benchmarked data from the published literature, and the results were compared with pre-trained models, including VGG16, ResNet50, VGG19, InceptionV3, and MobileNetV2. The experimental findings of the proposed method demonstrated a noteworthy classification accuracy of 97.84%, a precision success rate of 97.85%, a recall rate of 97.85%, and an F1-score of 97.90%. The results presented in this study showcase the exceptional accuracy of the proposed methodology in accurately classifying the most commonly occurring brain tumor types. The technique exhibited commendable generalization properties, rendering it a valuable asset in medicine for aiding physicians in making precise and proficient brain diagnoses.

Investigating UAV-Based Applications in Indoor-Outdoor Sports Stadiums and Open-Air Gatherings for Different Interference Conditions beyond 5G Networks.

With the onset of 5G technology, the number of users is increasing drastically. These increased numbers of users demand better service on the network. This study examines the millimeter wave bands working frequencies. Working in the millimeter wave band has the disadvantage of interference. This study aims to analyze the impact of different interference conditions on unmanned aerial vehicle use scenarios, such as open-air gatherings and indoor-outdoor sports stadiums. Performance analysis was carried out in terms of received power and path loss readings.

Massive MIMO NOMA: Double-Mode Model towards Green 5G Networks.

With the development of the Internet of Things (IoT), the number of devices will also increase tremendously. However, we need more wireless communication resources. It has been shown in the literature that non-orthogonal multiple access (NOMA) offers high multiplexing gains due to the simultaneous transfer of signals, and massive multiple-input-multiple-outputs (mMIMOs) offer high spectrum efficiency due to the high antenna gain and high multiplexing gains. Therefore, a downlink mMIMO NOMA cooperative system is considered in this paper. The users at the cell edge in 5G cellular system generally suffer from poor signal quality as they are far away from the BS and expend high battery power to decode the signals superimposed through NOMA. Thus, this paper uses a cooperative relay system and proposes the mMIMO NOMA double-mode model to reduce battery expenditure and increase the cell edge user's energy efficiency and sum rate. In the mMIMO NOMA double-mode model, two modes of operation are defined. Depending on the relay's battery level, these modes are chosen to utilize the system's energy efficiency. Comprehensive numerical results show the improvement in the proposed system's average sum rate and average energy efficiency compared with a conventional system. In a cooperative NOMA system, the base station (BS) transmits a signal to a relay, and the relay forwards the signal to a cluster of users. This cluster formation depends on the user positions and geographical restrictions concerning the relay equipment. Therefore, it is vital to form user clusters for efficient and simultaneous transmission. This paper also presents a novel method for efficient cluster formation.

A Data-Driven Intelligent Management Scheme for Digital Industrial Aquaculture based on Multi-object Deep Neural Network.

With the development of intelligent aquaculture, the aquaculture industry is gradually switching from traditional crude farming to an intelligent industrial model. Current aquaculture management mainly relies on manual observation, which cannot comprehensively perceive fish living conditions and water quality monitoring. Based on the current situation, this paper proposes a data-driven intelligent management scheme for digital industrial aquaculture based on multi-object deep neural network (Mo-DIA). Mo-IDA mainly includes two aspects of fish state management and environmental state management. In fish state management, the double hidden layer BP neural network is used to build a multi-objective prediction model, which can effectively predict the fish weight, oxygen consumption and feeding amount. In environmental state management, a multi-objective prediction model based on LSTM neural network was constructed using the temporal correlation of water quality data series collection to predict eight water quality attributes. Finally, extensive experiments were conducted on real datasets and the evaluation results well demonstrated the effectiveness and accuracy of the Mo-IDA proposed in this paper.

Game-o-Meta: Trusted Federated Learning Scheme for P2P Gaming Metaverse beyond 5G Networks.

The aim of the peer-to-peer (P2P) decentralized gaming industry has shifted towards realistic gaming environment (GE) support for game players (GPs). Recent innovations in the metaverse have motivated the gaming industry to look beyond augmented reality and virtual reality engines, which improve the reality of virtual game worlds. In gaming metaverses (GMs), GPs can play, socialize, and trade virtual objects in the GE. On game servers (GSs), the collected GM data are analyzed by artificial intelligence models to personalize the GE according to the GP. However, communication with GSs suffers from high-end latency, bandwidth concerns, and issues regarding the security and privacy of GP data, which pose a severe threat to the emerging GM landscape. Thus, we proposed a scheme, Game-o-Meta, that integrates federated learning in the GE, with GP data being trained on local devices only. We envisioned the GE over a sixth-generation tactile internet service to address the bandwidth and latency issues and assure real-time haptic control. In the GM, the GP's game tasks are collected and trained on the GS, and then a pre-trained model is downloaded by the GP, which is trained using local data. The proposed scheme was compared against traditional schemes based on parameters such as GP task offloading, GP avatar rendering latency, and GS availability. The results indicated the viability of the proposed scheme.

Innovative Spectrum Handoff Process Using a Machine Learning-Based Metaheuristic Algorithm.

A cognitive radio network (CRN) is an intelligent network that can detect unoccupied spectrum space without interfering with the primary user (PU). Spectrum scarcity arises due to the stable channel allocation, which the CRN handles. Spectrum handoff management is a critical problem that must be addressed in the CRN to ensure indefinite connection and profitable use of unallocated spectrum space for secondary users (SUs). Spectrum handoff (SHO) has some disadvantages, i.e., communication delay and power consumption. To overcome these drawbacks, a reduction in handoff should be a priority. This study proposes the use of dynamic spectrum access (DSA) to check for available channels for SU during handoff using a metaheuristic algorithm depending on machine learning. The simulation results show that the proposed "support vector machine-based red deer algorithm" (SVM-RDA) is resilient and has low complexity. The suggested algorithm's experimental setup offers several handoffs, unsuccessful handoffs, handoff delay, throughput, signal-to-noise ratio (SNR), SU bandwidth, and total spectrum bandwidth. This study provides an improved system performance during SHO. The inferred technique anticipates handoff delay and minimizes the handoff numbers. The results show that the recommended method is better at making predictions with fewer handoffs compared to the other three.

Efficient anomaly recognition using surveillance videos.

Smart surveillance is a difficult task that is gaining popularity due to its direct link to human safety. Today, many indoor and outdoor surveillance systems are in use at public places and smart cities. Because these systems are expensive to deploy, these are out of reach for the vast majority of the public and private sectors. Due to the lack of a precise definition of an anomaly, automated surveillance is a challenging task, especially when large amounts of data, such as 24/7 CCTV footage, must be processed. When implementing such systems in real-time environments, the high computational resource requirements for automated surveillance becomes a major bottleneck. Another challenge is to recognize anomalies accurately as achieving high accuracy while reducing computational cost is more challenging. To address these challenge, this research is based on the developing a system that is both efficient and cost effective. Although 3D convolutional neural networks have proven to be accurate, they are prohibitively expensive for practical use, particularly in real-time surveillance. In this article, we present two contributions: a resource-efficient framework for anomaly recognition problems and two-class and multi-class anomaly recognition on spatially augmented surveillance videos. This research aims to address the problem of computation overhead while maintaining recognition accuracy. The proposed Temporal based Anomaly Recognizer (TAR) framework combines a partial shift strategy with a 2D convolutional architecture-based model, namely MobileNetV2. Extensive experiments were carried out to evaluate the model's performance on the UCF Crime dataset, with MobileNetV2 as the baseline architecture; it achieved an accuracy of 88% which is 2.47% increased performance than available state-of-the-art. The proposed framework achieves 52.7% accuracy for multiclass anomaly recognition on the UCF Crime2Local dataset. The proposed model has been tested in real-time camera stream settings and can handle six streams simultaneously without the need for additional resources.

Stacked Model-Based Classification of Parkinson's Disease Patients Using Imaging Biomarker Data.

Parkinson's disease (PSD) is a neurological disorder of the brain where nigrostriatal integrity functions lead to motor and non-motor-based symptoms. Doctors can assess the patient based on the patient's history and symptoms; however, the symptoms are similar in various neurodegenerative diseases, such as progressive supranuclear palsy (PSP), multiple system atrophy-parkinsonian type (MSA), essential tremor, and Parkinson's tremor. Thus, sometimes it is difficult to identify a patient's disease based on his or her symptoms. To address the issue, we have used neuroimaging biomarkers to analyze dopamine deficiency in the brains of subjects. We generated the different patterns of dopamine levels inside the brain, which identified the severity of the disease and helped us to measure the disease progression of the patients. For the classification of the subjects, we used machine learning (ML) algorithms for a multivariate classification of the subjects using neuroimaging biomarkers data. In this paper, we propose a stacked machine learning (ML)-based classification model to identify the HC and PSD subjects. In this stacked model, meta learners can learn and combine the predictions from various ML algorithms, such as K-nearest neighbor (KNN), random forest algorithm (RFA), and Gaussian naive Bayes (GANB) to achieve a high performance model. The proposed model showed 92.5% accuracy, outperforming traditional schemes.

Blockchain-Based Peer-to-Peer Transactive Energy Management Scheme for Smart Grid System.

In Smart Grid (SG), Transactive Energy Management (TEM) is one of the most promising approaches to boost consumer participation in energy generation, energy management, and establishing decentralized energy market models using Peer-to-Peer (P2P). In P2P, a prosumer produces electric energy at their place using Renewable Energy Resources (RES) such as solar energy, wind energy, etc. Then, this generated energy is traded with consumers (who need the energy) in a nearby locality. P2P facilitates energy exchange in localized micro-energy markets of the TEM system. Such decentralized P2P energy management could cater to diverse prosumers and utility business models. However, the existing P2P approaches suffer from several issues such as single-point-of-failure, network bandwidth, scalability, trust, and security issues. To handle the aforementioned issues, this paper proposes a Decentralized and Transparent P2P Energy Trading (DT-P2PET) scheme using blockchain. The proposed DT-P2PET scheme aims to reduce the grid's energy generation and management burden while also increasing profit for both consumers and prosumers through a dynamic pricing mechanism. The DT-P2PET scheme uses Ethereum-blockchain-based Smart Contracts (SCs) and InterPlanetary File System (IPFS) for the P2P energy trading. Furthermore, a recommender mechanism is also introduced in this study to increase the number of prosumers. The Ethereum SCs are designed and deployed to perform P2P in real time in the proposed DT-P2PET scheme. The DT-P2PET scheme is evaluated based on the various parameters such as profit generation (for prosumer and consumer both), data storage cost, network bandwidth, and data transfer rate in contrast to the existing approaches.

GrAb: A Deep Learning-Based Data-Driven Analytics Scheme for Energy Theft Detection.

Integrating information and communication technology (ICT) and energy grid infrastructures introduces smart grids (SG) to simplify energy generation, transmission, and distribution. The ICT is embedded in selected parts of the grid network, which partially deploys SG and raises various issues such as energy losses, either technical or non-technical (i.e., energy theft). Therefore, energy theft detection plays a crucial role in reducing the energy generation burden on the SG and meeting the consumer demand for energy. Motivated by these facts, in this paper, we propose a deep learning (DL)-based energy theft detection scheme, referred to as GrAb, which uses a data-driven analytics approach. GrAb uses a DL-based long short-term memory (LSTM) model to predict the energy consumption using smart meter data. Then, a threshold calculator is used to calculate the energy consumption. Both the predicted energy consumption and the threshold value are passed to the support vector machine (SVM)-based classifier to categorize the energy losses into technical, non-technical (energy theft), and normal consumption. The proposed data-driven theft detection scheme identifies various forms of energy theft (e.g., smart meter data manipulation or clandestine connections). Experimental results show that the proposed scheme (GrAb) identifies energy theft more accurately compared to the state-of-the-art approaches.

Study of Multidimensional and High-Precision Height Model of Youth Based on Multilayer Perceptron.

Predicting the adult height of children accurately has great social value for the selection of outstanding athlete as well as early detection of children's growth disorders. Currently, the mainstream method used to predict adult height in China has three problems: its standards are not uniform; it is stale for current Chinese children; its accuracy is not satisfactory. This article uses the data collected by the Chinese Children and Adolescents' Physical Fitness and Growth Health Project in Zhejiang primary and secondary schools. We put forward a new multidimensional and high-precision youth growth curve prediction model, which is based on multilayer perceptron. First, this model uses multidimensional growth data of children as predictors and then utilizes multilayer perceptron to predict the children's adult height. Second, we find the Table of Height Standard Deviation of Chinese Children and fit the data of zero standard deviation to obtain the curve. This curve is regarded as Chinese children's mean growth curve. Third, we use the least-squares method and the mean curve to calculate the individual growth curve. Finally, the individual curve can be used to predict children's state height. Experimental results show that this adult height prediction model's accuracy (between 2 cm) of boys and girls reached 90.20% and 88.89% and the state height prediction accuracy reached 77.46% and 74.93%. Compared with Bayley-Pinneau, the adult height prediction is improved 19.61% for boys and 13.33% for girls. Compared with BoneXpert, the adult height prediction is improved 25.49% for boys and 6.67% for girls. Compared with the method based on the bone age growth map, the adult height prediction is improved 15.69% for boys and 24.45% for girls.