The Internet of Things (IoT) has become increasingly susceptible to cyber attacks, making it crucial to have strong detection mechanisms. This paper looks at using adversarial machine learning to boost IoT security. We implemented and evaluated Feedforward Neural Networks (FNN) and Long Short-Term Memory (LSTM) models on the Bot-IoT dataset for binary and multi-class classification tasks. We then evaluated how these models perform under Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD) adversarial attacks. Our results show high accuracy in detecting attacks under normal conditions. However, the models showed major weaknesses when faced with adversarial examples. This study highlights the urgent need for building adversarially robust machine learning models for IoT security. It also gives insights into how different model architectures perform against various attack intensities.

The growing scale and complexity of IoT (Internet of Things) networks demand robust security models that provide both secure identity authentication and reliable trust evaluation. While Self-Sovereign Identity (SSI) enables decentralized and user-centric authentication, ensuring the continuous trustworthiness of participating nodes remains a challenge. This paper presents a unified framework that combines SSI-based decentralized authentication with IOTA-ledger-assisted trust evaluation to provide an integrated security architecture for IoT systems. The proposed architecture guarantees that only authenticated nodes can participate in the network while continuously assessing their behavior and trustworthiness through immutable and tamper-resistant trust scores. Our approach is, to the best of our knowledge, the first to unify SSI-based authentication with IOTA-assisted trust evaluation, offering a holistic IoT security framework.

With internet connectivity's dynamic development and network communication escalating, an upward trend in advanced threats and cyber attack crimes is being witnessed, targeting sensitive data and critical systems. As the number of connected devices continues to increase, more opportunities are provided for malicious actors to exploit these vulnerabilities. Therefore, the attack surface expanded, particularly within the Internet of Things (IoT), reinforcing the necessity to design a suitable and adaptable security model for protecting data and minimizing the damage caused by network system intrusions and attacks. Machine learning is a swift and flexible way to develop a cybersecurity model. Clustering and other unsupervised learning techniques are frequently used to find hidden patterns or anomalies without needing labeled data. Still, they are limited in accurately distinguishing between normal and malicious behaviors, particularly when dealing with emerging attacks (such as zero-day attacks). This paper provides a hybrid semi-supervised classification-based clustering approach model to distinguish malicious emerging cyber-attacks. Initially, unsupervised clustering methods were implemented to group related data points and find underlying patterns to the reduced data dimensionality under Principal Component Analysis (PCA). Then, the pseudo-labels generated by clustering were used in supervised tasks with a Support Vector Machine (SVM) for classification. The experiments have been carried out on two IoT datasets, namely MQTTset and IoTID20, and the results confirm that the hybrid strategy is a viable and effective approach to dealing with IoT security issues in complicated and real-world situations.

The emergence of Software-defined Networks (SDN) has played a significant role in shaping the future of networking technologies. SDN aims to improve the flexibility, efficiency, and scalability of traditional networks by centralizing network control, allowing network administrators to manage and control the network through software applications. However, the flexibility provided by SDN architecture unveils numerous emerging network security concerns that require more attention to enhance SDN network security. So, in this paper, we propose a innovative intrusion detection system (IDS) for SDN using a hybrid model that combines CNN-BiLSTM and Transformer Encoder. The proposed approach was tested using the NSL-KDD and CICIDS2017 datasets, achieving accuracy rates of 99.7% and 99.68% respectively. The obtained results demonstrate that our proposed deep learning-based approach provides a robust security solution for detecting intrusions in SDN environments.

The Internet of Things (IoT) has revolutionized industries and daily life by connecting devices, sensors, and systems to autonomously collect, process, and share data. While offering immense possibilities, the distributed nature of IoT networks introduces significant security challenges. Conventional security measures struggle to address IoT's unique needs, and blockchain solutions face limitations. IOTA presents the Tangle, a novel Distributed Ledger Technology (DLT) tailored for IoT applications, offering feeless transactions, scalability, and decentralized consensus. This survey explores IOTA-based security solutions, focusing on access control, authentication, data integrity, confidentiality, threat modeling, risk assessment, performance, and scalability in IoT environments. Through a comprehensive review of literature, we highlight the efficacy of IOTA in enhancing IoT security. We identify future research directions, including interoperability, integration with emerging technologies, real-world deployments, and advanced trust management, to further strengthen IoT security using IOTA technology.

In the realm of future smart system development, one domain poised for significant growth is the emergence of smart cities worldwide, which aim to enhance people's quality of life. In fact, a smart city is characterized by the integration of a large volume of data and wireless communication enabled services, encompassing automated healthcare, smart transportation, home automation, smart parking, and traffic management, among others. However, the rapid growth of smart cities presents challenges in ensuring secure and reliable exchange of information between IoT devices and entities within the smart city ecosystem due to the enormous amount of data traffic generated by intelligent information systems. To address these challenges, the deployment of blockchain technology within smart cities holds promise for improving data integrity and facilitating the transparent, reliable, secure, and equitable delivery of services and applications. This paper explores the integration of Ethereum blockchain, smart city concepts, and the InterPlanetary File System (IPFS), investigating the potential for implementing an Ethereum blockchain-based smart contract on a smart city system. Finally, the proposed architecture is simulated in a limited environment, yielding results that demonstrate its feasibility.

Across the globe, agricultural yield faces numerous challenges, including unpredictable weather patterns, resource constraints, and the ever-present threat of plant diseases. Early and accurate disease detection is crucial for mitigating losses, optimizing resource allocation, and promoting sustainable farming practices. Machine Learning (ML) and Deep Learning (DL) techniques, particularly convolutional neural networks (CNNs), offer immense potential for tackling this challenge. This paper investigates the potential of DL for disease detection in agricultural crops. We delve into data scarcity, a major obstacle, and analyze the suitability of existing datasets like PlantVillage for training robust models. Furthermore, we explore popular CNN architectures, such as LeNet, AlexNet, and VGGNet, along with their strengths and limitations in the context of plant disease detection.

The Internet of Things (IoT) has grown exponentially over the past decade, but this growth has also raised a number of issues for the ongoing operation of IoT applications, including resource limitations, server overload, and the risk of improper use of private data. To address these challenges, Blockchain technology, which initially powered the crypto-currency Bitcoin, is gaining recognition as a solution that can enhance security and privacy. Blockchain (BC) provides several essential features, such as a consensus approach, peer-to-peer communications, trust without the need for a third party, and transactions controlled by conditions and functions through the use of smart contracts. Thus, BC technology is a suitable candidate for building a decentralized, autonomous Internet of Things system that addresses the issues raised by IoT. In this paper, we propose Secure Smart Home Communications based on the Ethereum BC and Smart Contract as a new design to solve the dilemma between the limited resources of IoT devices and the concerns of a centralized architecture. Quantitative and qualitative evaluations of the architecture under common threat models have highlighted its effectiveness in providing security and privacy for IoT applications.

The Software-Defined Networking (SDN) paradigm has introduced heightened flexibility and scalability to network infrastructure management. However, the centralized control plane inherent in SDN architectures is susceptible to an array of security vulnerabilities, necessitating the development of efficient and real-time anomaly detection systems. This paper presents a novel integrated methodology for real-time anomaly detection within SDN architectures, capitalizing on the synergies between Security Information and Event Management (SIEM) systems and advanced machine learning techniques to bolster network security. The proposed framework operates by seamlessly collecting and analyzing live network traffic data, promptly pinpointing potential anomalies, and subsequently correlating these events via the SIEM system. To enhance accuracy while mitigating false positives, machine learning algorithms are harnessed to accurately categorize network traffic into benign and malicious activities, dynamically adapting to evolving threat landscapes. Empirical validation is conducted through an exhaustive dataset of real-world network traffic, encompassing an extensive array of attack scenarios. Findings vividly underscore the efficacy of the amalgamated SIEM and machine learning-driven anomaly detection system, yielding impressive detection accuracy while maintaining notably low rates of false positives. Noteworthy is the system's intrinsic adaptability to emergent threats, culminating in an elevated caliber of network security and fortitude within the SDN domain. This contribution significantly enriches the realm of real-time anomaly detection research, endowing SDN architectures with a pioneering strategy to counteract intricate cyber threats effectively.

In 2020 only, 36 billion records have been leaked, 95% of those attacks have been caused by human error. Therefore organizations have been looking for multiple technologies to secure there system. One of the modern techniques is using machine learning for traffic classification which help us detect attacks based on monitoring data flow in our network or our workstation. In this paper we presents an implementation of new proposed model based on convolutional neural network (CNN) and long short term memory (LSTM). It is evident from the investigations that different machine learning methods can be used for intrusion detection. Further, the results demonstrated that the usage of machine learning techniques produce positive impact on improving the overall performance of the intrusion detection system in terms of accuracy and lowering false negatives. We using two of most known Deep Learning algorithms CNN and GRU (Convolutional Neural Network, Gated Recurrent Unit) to have a base ground regarding the performance of our proposed model. we are getting encouraging results with implementation of two of famous data-set (NSL-KDD and CIC-IDS2018) We tested our models on binary and multi-class classification for further observation.

Individuals and organizations face threats from attackers that constantly attempt to collect information by exploiting their vulnerabilities and obtain unauthorized access to the network. One way of protecting the system is by using an Intrusion detection System (IDS) alongside other security measures like firewalls, antivirus software, and access controls. An intrusion detection system, which is a system that can detect and recognize attacks on a network, is an approach that has been tried, tested, and tweaked to detect attacks as effectively as possible. Therefore, in this work, we will propose an approach using either a feature extraction method or a feature selection algorithm and other machine learning algorithms to provide a new way of intrusion detection, this approach will be applied to the NSL KDD data set.

This paper presents the findings of a field study conducted in Rabat, the capital of Morocco, utilizing the Wardriving technique to assess Wi-Fi network security. The study encompasses approximately 10,000 Wi-Fi networks situated in residential and administrative neighborhoods in Rabat. Through our comprehensive analysis, we observed that a substantial 89.42% of the networks use WPA2, suggesting that Wi-Fi security in Morocco compares favorably to that of developed countries. We also found that most networks don’t use default configurations, this indicates a proactive approach by network administrators to implement robust security measures. Moreover, our investigation revealed a balanced distribution of channels 1, 6, and 11, illustrating that network operators are mindful of potential interferences, particularly on channel 6, and have taken measures to mitigate such interferences effectively. Based on our results, we draw a positive conclusion that the Wi-Fi situation in the examined neighborhoods of Rabat is highly encouraging. The study highlights the efforts made by network administrators to secure their Wi-Fi infrastructures and optimize network performance, contributing to a safer and more reliable wireless environment for users. This research serves as a valuable reference for understanding the state of Wi-Fi security in Rabat and provides insights into the overall Wi-Fi landscape in the city. The data-driven conclusions can aid policymakers, businesses, and individuals in further enhancing Wi-Fi security practices to ensure the continued growth and stability of wireless connectivity in the region.

With the rapid increase in network intrusions, applying an active network intrusion defense is more important than ever before. Different learning algorithms have been combined to achieve better performance. To improve the accuracy and efficiency of the network intrusion detection system, a new hybrid algorithm is designed, which combines the Random forest, Decision tree and Multilayer Perceptron (MLP) algorithms. The experimental results show that the hyprid model has a better true Positive Rate (TPR) for attack activities, rapidly data preprocessing speed, and shorter training time. In particular, the accuracy of multi-class classification can reach as high as 99.7% in the NSL-KDD dataset, 77.99% in the UNSW-NB15 dataset and 84.89% in the CIC-IDS-2017.

In the Internet of Things (IoT), physical objects are connected to the Internet and can collect, transmit, and receive data. Advances in microelectronics technology have led to smaller and more diverse IoT devices with increased capabilities. Cloud and edge technologies are also providing these devices with efficient computing, storage, and low-latency solutions. This has led to the widespread use of IoT in a variety of areas, but it also presents an increased risk of cyberattacks, data breaches, and unauthorized access to sensitive information. However, studies and analyses of physical object security and privacy in the IoT focus on identifying and mitigating these risks. This paper studied the challenges of IoT systems and different security requirements. On the other hand, the mapping of attack methods, architecture, and security measures of physical components in IoT systems Privacy, security, and architecture requirements will be associated with physical components based on their life cycles.

In the last few years, the Internet Of Things (IoT) technology has become more and more used and widespread in different aspects of daily life thanks to intelligent services offered in several fields (robotics, home, hospitals, etc). IoT systems are vulnerable to a variety of security attacks, and they are facing multiple risks and considered targets for cyberattacks, for this reason, securing IoT system is a major challenge, existing security protocols based on tradition are not suitable for them. To cope with different security challenges, Machine Learning (ML) Techniques are able to provide intelligence for IoT devices and networks. This paper presents the security problems and the existing ML solutions to manage security aspects related to the IoT domain. This paper proposes a classification model to detect attack on the UNSW-NB18 dataset and implements the following algorithms namely, Decision Tree (DT), Random Forest (RF), Logistic Regression (LR) and k-Nearest Neighbors (KNN). The best results were achieved by the Random Forest algorithm, with an accuracy of 99.96%.

Nowadays, Internet of Things (IoT) devices occupy a large part of our lives as consumers of these technologies or potential ones. Yet, they present significant privacy and security risks. These risks play a significant role in the potential consumers' decision-making process regarding the use of these smart devices along with other parameters such as price, availability, ergonomics etc. To understand how perceived risk affects the intention of potential consumers to use connected objects, research based on models combining IT security and data privacy on one hand and social and behavioral parameters on the other have been conducted. The results of these studies showed that perceived security and privacy risks have real impact on potential consumers' intention to use IoT devices. In this paper we have performed a review of the results of the conducted studies and highlighted the common findings regarding the impact of security and privacy risks. The results of this paper represent a state of the art of the main work done to date on this subject and form a foundation on which future studies can be based.

In the last ten years, many variants of the principal component analysis were suggested to fight against the curse of dimensionality. Recently, A. Sharma et al. have proposed a stable numerical algorithm based on Householder QR decomposition (HQR) called QR PCA. This approach improves the performance of the PCA algorithm via a singular value decomposition (SVD) in terms of computation complexity. In this paper, we propose a new algorithm called RRQR PCA in order to enhance the QR PCA performance by exploiting the Rank-Revealing QR Factorization (RRQR). We have also improved the recognition rate of RRQR PCA by developing a nonlinear extension of RRQR PCA. In addition, a new robust RBF Lp-norm kernel is proposed in order to reduce the effect of outliers and noises. Extensive experiments on two well-known standard face databases which are ORL and FERET prove that the proposed algorithm is more robust than conventional PCA, 2DPCA, PCA-L1, WTPCA-L1, LDA, and 2DLDA in terms of face recognition accuracy.

It is well-known that intrusion detection systems are an effective way to detect malicious connections in a computer network. Different feature extraction techniques have been applied in the field of detection intrusions; the most common one is Principal Component Analysis (PCA). Nevertheless, PCA is restricted to linear principal components and suffers from sensitivity to noise and can be easily affected by outliers. To deal with the drawbacks of PCA many data dimensionality reduction methods have been proposed, among them we found Robust Fuzzy PCA (RFPCA) which employs fuzzy memberships in order to reduce the effect of outliers. Unfortunately, even though RFPCA showed satisfactory results, it still suffers from the influence of outliers. In fact, using an increasing function such as exponential function in the estimation of memberships will assign a big membership values to outliers, consequently, the obtained results can be skewed. In this paper, we suggest a new variant of Robust Fuzzy PCA (RFPCA) method for the purpose of network IDS. Extensive experiments on the two well known datasets i.e. KDDcup99 and NSL-KDD demonstrated that the proposed approach takes the advantage over RFPCA and PCA in terms of network attacks detection and false alarms reduction.

Countering network threats, particularly intrusions, is a challenging area of research in the field of information security. Intruders use sophisticated mechanisms to hide the attack payload and break the detection techniques. To overcome that, various unsupervised learning approaches from the field of machine learning and pattern recognition have been employed. The most popularly used method is Principal Component Analysis (PCA). It proposes to extract the critical features of a network connection, then, it exploit them to identify the intrusion. However, PCA approach is prone to outliers due to the square G-norm based objective function. As a solution to that, many PCA variants such R1-PCA and G-PCA were proposed. Nevertheless, They still work with the square l2-norm distance based mean, which is not the optimal mean. This paper introduces a new variant of PCA namely QR-OMPCA. Firstly, this method integrates the mean calculation into the feature extraction function, such that the optimal mean can be obtained to enhance the intrusion detection accuracy. Secondly, it incorporates a fast QR decomposition. Experiments on KDDcup99 and NSL-KDD datasets confirm the superiority of the proposed method over many PCA variants in terms of intrusion detection accuracy and CPU time reduction.

The rapid evolution of information and communication technologies leads to a big networks security problem. For this reason, the Intrusion Detection System (IDS) has been developed in order to detect and prevent computer network attacks. However, the majority of IDSs operate on huge network traffic data with many useless and redundant features. Consequently, the IDS generates a lot of false alarms and the intrusion detection process becomes difficult and imprecise. To improve the performance of an IDS, many data dimensionality reduction methods, such as Principal Component Analysis (PCA), have been proposed. However, the classical PCA approach, that is based on the covariance matrix of the data, is very sensitive to outliers. In order to overcome this problem, we propose to introduce a new variant of PCA namely L1-norm PCA. This new method is based on the L1-norm maximization, which is more robust to outliers, instead of the Euclidean norm in the classical PCA. Extensive experiments on the well-known KDDcup99 dataset are exploited for testing the effectiveness of the proposed approach. Obtained results confirm the superiority of L1-norm PCA over the traditional PCA in terms of network attacks detection and false alarms reduction.

This paper presents two novels algorithms based on fuzzy logic and discriminant analysis for face recognition. The first one is a fuzzy extension of a linear discriminant analysis algorithm namely LDA/QR and the second one is a fuzzy extension of kernel scatter-difference based discriminant analysi s (KSDA) algorithm. They can deal with small sample size and nonlinear problems which degrade the performance of face recognition system. Experimental results on the ORL and the extended Yale B face databases show that the two proposed approaches, using fuzzy logic, achieves a better performance in face recognition compared with their original versions.