We leverage realistic datasets and evaluate online learning strategies for downlink beam-pair selection in mmWave networks, including classical contextual-bandit methods, deep reinforcement learning approaches, and our proposed two-stage approach, termed warm-start Neural-Linear Thompson Sampling (NLTS). Unlike standard methods that learn solely online, our approach pre-trains a neural network offline using past channel measurements and signal strengths, then performs lightweight Bayesian updates in the learned feature space during deployment. Experiments show that our proposed warm-start NLTS significantly accelerates learning, improves beam-pair selection accuracy, and enhances spectral efficiency outperforming all baselines with only a modest increase in the decision latency. These findings highlight the potential of combining supervised pre-training with online bandits for faster, more reliable beam-forming in next-generation wireless networks.

The rapid growth of internet technologies, combined with the robust and evolving infrastructure of the Internet of Things (IoT) networks, has led to the presence of a variety of sophisticated attacks, highlighting the need for advanced defensive measures. Artificial Intelligence (AI) and Machine Learning (ML) have become key tools to tackle a spectrum of attack vectors, as they deliver intelligent, adaptive, context-driven, and domain-specific responses. This work attempts to provide a comprehensive analysis of emerging IoT attack identification in real time by proposing an ensemble framework that incorporates two balancing strategies: Synthetic Minority Oversampling Technique (SMOTE) and Class Weights adjustment. More specifically, the framework consists of three main stages, including data collection from the CIC IoT-DIAD 2024 benchmark, data preprocessing, and flow-based feature selection. The next immediate step addresses label imbalance using SMOTE, which oversamples minority cases by generating synthetic data points, and class weights, which adjust the loss function to apply greater penalties for misclassifying minority instances. Following this, ensemble models are developed and optimized using bagging and stacking meta-learners, with hyperparameters tuned through Optuna. Models evaluation demonstrates improved performance across valuable metrics (Accuracy(0.96-0.97), Precision(0.94-0.99), Recall(0.94-0.99), F1-score(0.96-0.97) and AUC(0.99)) when the models sensitivity to the classes is adjusted, rather than relying solely on synthetic data generation. These results showcase the practical evidence of the proposed framework in enhancing the security and trust of real IoT systems, further supported by a low false alarm rate(0.000339) and false discovery rate(0.000326).

Machine learning (ML)-based network anomaly detection methods are proven to provide automated network protection from traffic misbehavior and authorized system access through data monitoring and analysis. However, conventional centralized methods present risks for data privacy and breaches. By facilitating distributed model training over a number of network nodes, Federated Learning (FL) emerges as a key enabler for effective anomaly detection yet while preserving the privacy of the data. This paper studies FL-based detection approaches under two different Deep Learning models, CNN and MLP. We use XGBoost for feature selection and the two UNSWNB15 and CICDDoS2019 datasets for assessing the effectiveness of each model through the evaluation of standard performance metric criteria, namely the recall, precision, accuracy, and F1score metrics. Our experimental findings indicate that integrating XGBoost-based feature selection with the CNN model yields superior performance on the UNSW-NB15 dataset, whereas the MLP model benefits more from the same integration when applied to the CICDDoS2019 dataset.

Detecting irregular patterns that suggest potential threats or system flaws plays a vital role in anomaly detection, which is essential for maintaining the security and integrity of network systems, especially in IoT environments where devices are often vulnerable and widely distributed. This study uses the UNSW-NB15 dataset, a vast collection of network traffic data, to examine several Machine Learning (ML) and Deep Learning (DL) approaches for network anomaly detection. In order to improve model performances, the study uses strategies such the Synthetic Minority Over-Sampling Technique (SMOTE) to address the issue of class imbalance. The capabilities of a number of widely used machine learning (ML) algorithms, such as Decision Trees, Random Forests, KNN, XGB, and widely used deep learning (DL) models, such as CNN, ANN and LSTM, to identify unusual patterns in various attack and typical behavior scenarios are assessed. Our findings provide important information for future advancements in anomaly detection techniques by highlighting the significance of feature selection, class balancing, and model resilience in successfully differentiating unusual behaviors in network traffic. XGB emerged as the most successful approach in this study, with its enhanced performance largely attributed to the synergy between its robust ensemble framework and the class balancing achieved through SMOTE.

Rising phishing attacks pose serious cybersecurity threats due to their use of fraudulent links to collect confidential user information. In this paper, we evaluate the performance of various Machine Learning (ML) models, including Decision Trees, Random Forest, and Extreme Gradient Boosting, to address this growing threat. Additionally, we assess the effectiveness of different feature selection techniques, such as Analysis of Variance, Correlation-based Selection, Mutual Information, and Recursive Feature Elimination with Cross-Validation. Our findings demonstrate that combining Extreme Gradient Boosting with Recursive Feature Elimination and Cross-Validation outperforms previous methods. The proposed solution achieved an accuracy of 97.33%, a recall of 97.1656%, an F1 score of 97.3%, and a precision of 97.42%, highlighting its potential for effectively identifying phishing attacks

In several industrial applications, visual inspection impacts production and quality of products. In case of surface defect on metal structures, a decision should be made to discard the defective piece and prevent from possible failure in further steps of an industrial process. Accordingly, the decision-making should be automated using advanced image processing methods which must take into consideration the various surface defect types. Nevertheless, they can sometimes look similar causing a high level of ambiguity, even from the point of view of an expert. Possibility theory can offer a formalism and allows coping with such ambiguity, yielding more reliable decisions. Moreover, possibility transform based on probability intervals can tackle the problem of lack of information or sampling problem, and can describe in a more realistic way the features within a specific confidence region level based on Goodman concept. Then, a further transformation of interval probability to possibility using Masson et Denoeux formalism, allows a possibility mapping of the different classes at a given confidence level. In this paper, we propose a new formalism based on possibility transform, handling probability intervals, for addressing ambiguity and uncertainty affecting steel surface defects. The effectiveness of the proposed method is largely confirmed by carrying out comparative studies with other methods similar to deep learning or conventional methods. Accordingly, two validation databases NEU-DET and GC10-DET are deployed, demonstrating that the the possibility approach allows an improvement in the classification of surface defects by achieving accuracy rates sometimes approaching 100%, outperforming for the majority of the considered defect types, existing approaches.

Researchers are turning to advanced networks and technology to help cities manage resource constraints. With this continuing development in cities, smart mobility applications, that mitigate problems in citizen's life, are anticipated to increase exponentially in the coming period. Connected computers and mobiles are attacked by software robots and spiders acting as human beings to gain access. Various systems, referred to as CAPTCHA (Completely Automated Public Turing test to tell Computers and Human Apart), are widely employed to leverage the aesthetic complexity of Arabic script while effectively thwarting automated attacks. This paper introduces a method aimed at differentiating Arabic-speaking users from computer programs using Arabic texts. We present novel methods and techniques for generating captchas with Arabic handwritten calligraphy, focusing on balancing readability and security. Through extensive testing and evaluation, we demonstrate the effectiveness of our approach in resisting automated attacks while ensuring a seamless user experience. Our findings underscore the potential of leveraging Arabic handwritten calligraphy in captcha generation to enhance online security, offering a promising avenue for mitigating automated threats in the digital landscape.

ChatGPT has been acknowledged as a powerful tool that can radically boost productivity across a wide range of industries. It reveals potential in cybersecurity-related tasks such as social engineering. Nevertheless, this possibility raises important concerns regarding the thin line separating moral use of this technology from its harmful usage. It is imperative to address the challenges of distinguishing between legitimate and malevolent use of ChatGPT. This research paper investigates the many concerns of ChatGPT in cybersecurity, privacy and enterprise settings. It covers harmful attacker uses such as injecting malicious prompts, testing brute force attacks, preparing and developing ransomware attacks, etc. Defenders’ proactive activities are also addressed, highlighting ChatGPT’s significance in security operations and threat intelligence. These defensive operations are classified based on the National Institute of Standards and Technology cybersecurity framework. They involve analyzing configuration files, inquiring about authoritative server, improving security in various systems, etc. Moreover, secure enterprise practices and mitigations spread through five classes are proposed, with an emphasis on clear usage standards and guidelines establishment, personally identifiable information protection, adversarial attack prevention, watermarking generated content, etc. An integrated discussion digs into the interaction of offensive and defensive applications, covering ethical and practical concerns. Future attacks are also discussed, along with potential solutions such as content filtering and collaboration. Finally, a comparative analysis with recent research on ChatGPT security concerns is directed. The paper provides a thorough framework to comprehend the range of implications associated with ChatGPT, enabling the navigation of cybersecurity and privacy challenges.

The fast spread of Covid-19 or the novel Coronavirus in the world has influenced it and caused a huge number of deaths. This remains a disastrous warning to general wellbeing and will be set apart as probably the most dangerous pandemic in world history and one of the important health challenges that the world has ever faced. The public health policymakers need the dependable forecasting of the active cases of Covid-19 to plan the future medical facilities. In this work, Machine Learning has been used to forecast the number of active cases of Covid-19 in some countries and in the world using John Hopkins University's data to track the outbreak, attached by Desktop and Web application using Tkinter and Flask, python's frameworks for visualizing the data in the affected countries that gives an understandable form of the data powered by different types of charts and choropleth maps and predictions of active cases of Covid-19 which have brought suffering to people everywhere based on two Models (ARIMA and Polynomial Regression).

Urban Traffic Networks are characterized by their high dynamics and increased traffic congestion cases, leading to a more complex road traffic management. The present research work suggests an innovative advanced vehicle guidance system based on Hierarchical Interval Type-2 Fuzzy Logic model optimized by the Particle Swarm Optimization (PSO) method. Indeed, this system allows an intelligent and prompt adjustment of the road traffic network in a dynamic way and improves the entire road network quality, particularly in case of congestions or jams, considering real-time traffic information. The best followed road is selected according to the quality of traffic and route length, together with contextual factors pertaining to the driver, the environment, and the path. The proposed system is executed and simulated using SUMO (Simulation of Urban Mobility), for which four large areas situated in the cities of Sfax, Luxembourg, Bologna and Cologne have been tested. The simulation results proved the effectiveness of learning the Hierarchical Interval Type-2 Fuzzy Logic model using PSO real time technique to accomplish multi-objective optimality regarding two criteria: number of cars that attain their destination and average travel time. The obtained results have confirmed the efficiency of the proposed system.

Traffic congestion affects quality of life by inducing frustration and wasting time. The congestion is also critical to vehicles with high emergencies such as ambulances or police cars. This leads to additional CO2 emissions. Traffic management requires the accurate modeling of congestion levels. Two main observable parameters identify the congestion state of a city: vehicle speed and density. Congestion has an intuitive definition rather than a quantitative one, and is associated with the disorder and randomness occurring in traffic parameters. Therefore, statistical analysis offers an efficient and natural framework for modeling such disorders. In this study, a differential-entropy-based approach was proposed for labelling purposes. Subsequently, supervised congestion prediction from traffic meta-parameters based on a convolutional neural network was proposed. Traffic parameters includes node localization, date, day of the week, time of day, special road conditions, and holidays. The proposed model is validated on the CityPulse dataset, which is a set of vehicle traffic records, collected in Aarhus city in Denmark over a period of six months, for 449 observation nodes. Simulation results on the CityPulse dataset illustrate that the proposed approach yields accurate prediction rates for different nodes considered. The proposed system can prevent traffic congestion by reorienting the drivers to follow other itineraries.

Traffic congestion is a problem in most large cities world wide. It occurs when the capacity of road is surpassed, resulting in augmented vehicular queuing and slower average speeds. The traffic congestion can be caused or increased by various conditions like weather, road work, road traffic incidents. To deal with these problems, we propose a novel cooperative Multi-Agent system (MAS) for Road Traffic Decision Making in Vehicular Ad-Hoc network (VANET) based on a Hierarchical Interval Type-2 Fuzzy Knowledge Representation System (CMRHFS) used for travel route guidance. Our proposal aims to increase the road safety and the quality of the entire road network, especially in case of congestions, accidents and jams, considering traffic information in real-time as well as drivers travel time to attain their destinations. The obtained simulation results have proved our suggested system efficiency compared to Dijkstra's algorithm and Hierarchical Interval Type-2 Fuzzy Logic System (HIT2FLS) regarding two criteria: average travel time and path flow.

In this work, we address the problem of logistic management of Implantable Medical Devices (IMD) at hospital centers in Tunisia. Our goal is to propose a radio frequency identification (RFID) based framework to secure and automate the management of IMD between the pharmacy and the care units. The RFID technology has been widely used in various fields, including the health care sector, to quickly identify moving and/or remote objects. Yet, to date there is no recognized RFID based system appropriate for the purpose of IMD management. For this reason and due to the pressing need to improve safety and traceability in public health care environment, we propose a tailored solution to adapt RFID technology to the logistic management of IMD at La Rabta university health centre which is the prominent cornerstone of the health sector in Tunisia. This solution may be considered as a prototype that might be extended and deployed either inside or outside the region of Tunis.

Traffic congestion leads to many problems, namely road users' dissatisfaction, air pollution and waste of time and fuel. For this reason, congestion detection at an early stage is required to perform an efficient exploitation of resources. This paper proposed a Hierarchical Type-2 Beta Fuzzy Knowledge Representation system for the selection of optimal route. Consequently, this system aims to avoid longer travel times, and to decrease traffic accidents and the number of traffic congestion situations. The selection is performed through itineraries assessment by contextual factors such as Max speed and density of a given path. For the validation, the traffic simulation was done with the open source microscopic road traffic simulator SUMO. When compared with the Dijkstra's algorithm, the proposed system showed better performance in terms of average travel time and path flow. These promising results prove the potential of our method to relieve traffic congestion.

Smart City automation has become crucial notion for improving the quality of the citizens' lives, which gives growth to smart cities. Such improvement requires an increase in the number of Electronic Control Units (ECUs) and networking technologies in the automotive field and, consequently in the amount of exchanged signals and messages on the in-vehicle network. This upgrading imposed new requirements in terms of data rates, performance, capacity, efficiency, etc. A specialized traffic analyzing system is well needed to scrutinize all exchanged data. These systems allow automotive network diagnostic and ECU's testing by controlling local networks and giving a graphic display of network traffic statistics. This paper presents the challenges of a high performance in-car network traffic analyzer and the use of the analysis results as input data in a type 2 fuzzy Rule base system for road choice.

Numerous electronic systems have been added to the vehicle to provide more comfort and safety to the drivers. This has led to a rise in the number of ECUs (Electronic Control Units) and consequently an eminence in the number of connections and networks as well as the amount of transmitted information. With these innovations and advances in technology and their high requirements, Ethernet is introduced to the automotive industry to profit from its benefits. Among the applications that this protocol has occupied we found the analysis and diagnosis of network traffic flowing between the ECU of the car. To successfully analyze all the exchanged data between the ECUs, an efficient analysis system is required. Indeed, this system should optimally capture all Ethernet packets without any loss allowing the realization of filters, obtaining statistics and real-time display.

The road traffic becomes more complex to manage because of the high dynamics of traffic flow and the rise of travel time when the number of vehicles augments in the road networks. Hence, the shortest itinerary based on route length (as provided by GPS navigators) cannot be the best solution nowadays. The application of type-2 Fuzzy Logic is regarded as an effective way for transportation engineering to prevent the problem of ambiguity and uncertainty of road perceptions. In this paper, we propose a hierarchical type-2 Fuzzy Logic System to evaluate itinerary by integrating contextual factors influencing the route choice like speed and road work information.

Usually, road networks are characterized by their great dynamics including different entities in interactions. This leads to more complex road traffic management. This paper proposes an adaptive multiagent system based on the ant colony behavior and the hierarchical fuzzy model. This system allows adjusting efficiently the road traffic according to the real-time changes in road networks by the integration of an adaptive vehicle route guidance system. The proposed system is implemented and simulated under a multiagent platform in order to discuss the improvement of the global road traffic quality in terms of time, fluidity and adaptivity. (C) 2014 Elsevier Ltd. All rights reserved.