Comparative Analysis of LSTM-Based Variant Models for Detecting Attacks in IoT Networks

Background. With the proliferation of Internet of Things (IoT) networks in the digital age, the risk of cyberattacks, especially DDoS attacks, is also increasing. IoT devices such as smart refrigerators, thermostats, webcams and other Internet-connected home appliances are being targeted by attackers who can use them as part of a botnet to conduct DDoS attacks. These devices often have inadequate network security and are rarely updated, making them vulnerable. DDoS attacks can result in significant losses such as lost revenue, reputational damage and costs to restore services. So, the vulnerability of IoT networks to DDoS attacks and the need to develop effective protection measures is a pressing issue. Objective. The purpose of the paper is to analyse software that supports DDoS attacks in IoT networks. Provide general recommendations to help improve approaches to defence measures in IoT networks against DDoS-enabled malware. Methods. Five main DDoS attack models are considered: agent-handler model, reflexive model, IRC-based model, web-based model, and P2P-based model. Three most dangerous software capable of DDoS attacks on IoT networks are analyzed: Mirai, XOR.DDoS and Linux.Hydra. Results. There are many models and methods of DDoS attacks on IoT networks. The most dangerous are Mirai, XOR.DDoS and Linux.Hydra. Each of these software has its own specific characteristics and methods of carrying out attacks. The study also showed that there are several effective measures to counter these attacks, including setting strong passwords, regularly updating software, setting up traffic filters and restricting network access. Conclusions. Key aspects of DDoS attacks, their models and process steps are considered. The paper focuses on the three most dangerous software used to conduct such attacks and provides recommendations on how to counteract them.

The Internet of Things (IoT) presents great potential in various fields such as home automation, healthcare, and industry, among others, but its infrastructure, the use of open source code, and lack of software updates make it vulnerable to cyberattacks that can compromise access to data and services, thus making it an attractive target for hackers. The complexity of cyberattacks has increased, posing a greater threat to public and private organizations. This study evaluated the performance of deep learning models for classifying cybersecurity attacks in IoT networks, using the CICIoT2023 dataset. Three architectures based on DNN, LSTM, and CNN were compared, highlighting their differences in layers and activation functions. The results show that the CNN architecture outperformed the others in accuracy and computational efficiency, with an accuracy rate of 99.10% for multiclass classification and 99.40% for binary classification. The importance of data standardization and proper hyperparameter selection is emphasized. These results demonstrate that the CNN-based model emerges as a promising option for detecting cyber threats in IoT environments, supporting the relevance of deep learning in IoT network security.

A cascaded federated deep learning based framework for detecting wormhole attacks in IoT networks

With the increasing popularity of the Internet of Things (IoT) platforms, the cyber security of these platforms is a highly active area of research. One key technology underpinning smart IoT systems is machine learning, which classifies and predicts events from large-scale data in IoT networks. Machine learning is susceptible to cyber attacks, particularly data poisoning attacks that inject false data when training machine learning models. Data poisoning attacks degrade the performances of machine learning models. It is an ongoing research challenge to develop trustworthy machine learning models resilient and sustainable against data poisoning attacks in IoT networks. We studied the effects of data poisoning attacks on machine learning models, including the gradient boosting machine, random forest, naive Bayes, and feed-forward deep learning, to determine the levels to which the models should be trusted and said to be reliable in real-world IoT settings. In the training phase, a label modification function is developed to manipulate legitimate input classes. The function is employed at data poisoning rates of 5%, 10%, 20%, and 30% that allow the comparison of the poisoned models and display their performance degradations. The machine learning models have been evaluated using the ToN_IoT and UNSW NB-15 datasets, as they include a wide variety of recent legitimate and attack vectors. The experimental results revealed that the models’ performances will be degraded, in terms of accuracy and detection rates, if the number of the trained normal observations is not significantly larger than the poisoned data. At the rate of data poisoning of 30% or greater on input data, machine learning performances are significantly degraded.

Destructive cyber-attacks and cybercriminals are increasing with the increase in IoT (Internet of Things) devices globally. This ha/s led to the need for increase in security in IoT systems. Innovative and novel techniques are used by the intruders to accomplish malicious goals effectively through cyber-attacks. An Intrusion Detection System (IDS) is used for classification of attacks in IoT networks based on anomaly detection and machine learning techniques. Inefficiency is observed in the conventional machine learning models and intrusion detection techniques as the network technologies are unpredictable. Accurate identification of various anomalies is possible with deep learning models in several research segments. The input data along with its prominent characteristics may be categorized automatically for classification and anomaly detection using convolutional neural networks (CNN). Faster computations are enabled due to the performance efficiency of CNN. For IoT networks, an intrusion detection model based on anomaly detection is designed and developed in this paper. A multiclass classification framework is created initially using a CNN model. Further, 3D CNN is used for implementation of the proposed model. Various intrusion detection datasets from IoT networks are used for validation of the proposed CNN model. Pre-trained multiclass CNN model is used for implementation of multiclass and binary classification based on transfer learning. When compared to the conventional deep learning models, the proposed multiclass and binary classification framework has attained improved F1 score, recall, precision and accuracy.

The exponential growth of the Internet of Things (IoT) has led to the rapid expansion of interconnected systems, which has also increased the vulnerability of IoT devices to security threats such as distributed denial-of-service (DDoS) attacks. In this paper, we propose a machine learning pipeline that specifically addresses the issue of DDoS attack detection in IoT networks. Our approach comprises of (i) a processing module to prepare the data for further analysis, (ii) a dynamic attribute selection module that selects the most adaptive and productive features and reduces the training time, and (iii) a classification module to detect DDoS attacks. We evaluate the effectiveness of our approach using the CICI-IDS-2018 dataset and five powerful yet simple machine learning classifiers—Decision Tree (DT), Gaussian Naive Bayes, Logistic Regression (LR), K-Nearest Neighbor (KNN), and Random Forest (RF). Our results demonstrate that DT outperforms its counterparts and achieves up to 99.98% accuracy in just 0.18 s of CPU time. Our approach is simple, lightweight, and accurate for detecting DDoS attacks in IoT networks.

The growing development of IoT (Internet of Things) devices creates a large attack surface for cybercriminals to conduct potentially more destructive cyberattacks; as a result, the security industry has seen an exponential increase in cyber-attacks. Many of these attacks have effectively accomplished their malicious goals because intruders conduct cyber-attacks using novel and innovative techniques. An anomaly-based IDS (Intrusion Detection System) uses machine learning techniques to detect and classify attacks in IoT networks. In the presence of unpredictable network technologies and various intrusion methods, traditional machine learning techniques appear inefficient. In many research areas, deep learning methods have shown their ability to identify anomalies accurately. Convolutional neural networks are an excellent alternative for anomaly detection and classification due to their ability to automatically categorize main characteristics in input data and their effectiveness in performing faster computations. In this paper, we design and develop a novel anomaly-based intrusion detection model for IoT networks. First, a convolutional neural network model is used to create a multiclass classification model. The proposed model is then implemented using convolutional neural networks in 1D, 2D, and 3D. The proposed convolutional neural network model is validated using the BoT-IoT, IoT Network Intrusion, MQTT-IoT-IDS2020, and IoT-23 intrusion detection datasets. Transfer learning is used to implement binary and multiclass classification using a convolutional neural network multiclass pre-trained model. Our proposed binary and multiclass classification models have achieved high accuracy, precision, recall, and F1 score compared to existing deep learning implementations.

With the emergence of the Internet of Things (IoT) and the increasing number of resource-constrained interconnected smart devices, there is a noticeable increase in the number of cyber security crimes. In the face of the possible attacks on IoT networks such as network intrusion, denial of service, spoofing and so on, there is a need to develop efficient methods to locate vulnerabilities and mitigate attacks in IoT networks. Without loss of generality, we consider only intrusion-related threats to IoT. A honeypot is a system used to understand the potential dynamic threats and act as a proactive measure to detect any intrusion into the network. It is used as a trap for intruders to control unauthorized access to the network by analyzing malicious traffic. However, a sophisticated attacker can detect the presence of a honeypot and abort the intrusion mission. Therefore it is essential for honeypots to be undetectable. In this paper, we study and analyze possible techniques for SSH and telnet honeypot detection. Moreover, we propose a new methodology for probabilistic estimation of honeypot detection and an automated software implemented this methodology.

With the widespread of cyber attacks, network intrusion detection system (NIDS) is becoming an important and essential tool to protect Internet of Things (IoT) environments. However, it is well-known that the NIDS performance depends heavily on the effectiveness of detection model, which can be influenced significantly by the learning mechanism and the available training data. Many existing studies try to mitigate the above challenges, but few of them consider the adaptability and the cost of deploying an NIDS, the integrity of learning process, the capacity of model based on concrete traffic samples at the same time. To fill this gap and improve the detection performance, we propose a collaborative learning based detection framework called ADCL, which can mitigate the limitations on the knowledge of a single model by leveraging multiple models trained in similar environments and detecting intrusions in a collaborative manner. Our evaluation results indicate that ADCL can provide better performance compared with a single model on detecting various attacks in IoT networks. Specifically, ADCL improves F-score by up to 80% for adaptability, 42% in mitigating the reliance on learning integrity, 85% for model capacity. Furthermore, the detection results of ADCL guide those single models to update and increase the F-score by 15%.

Permissionless distributed ledgers provide a promising approach to deal with the Internet of Things (IoT) paradigm. Since IoT devices mostly generate data transactions and micropayments, distributed ledgers that use fees to regulate the network access are not an optimal choice. In this paper, we study a feeless architecture developed by IOTA and designed specifically for the IoT. Due to the lack of fees, malicious nodes can exploit this feature to generate an unbounded number of transactions and perform a denial of service attacks. We propose to mitigate these attacks through verifiable delay functions. These functions, which are non-parallelizable, hard to compute, and easy to verify, have been formulated only recently. In our work, we design a denial of service prevention mechanism which addresses network heterogeneity, limited node computational capabilities, and hardware-specific implementation optimizations. Verifiable delay functions have mostly been studied from a theoretical point of view, but little has been done in tangible applications. Hence, this paper can be considered as a pioneer work in the field, since it builds a bridge between this theoretical mathematical framework and a real-world problem.

Securing Internet of Things (IoT) against attacks is a very interesting area of research. A cyberattack refers to as any form of malicious activity that targets IT systems, networks and/or people with a view to gaining illegal access to systems and data they contain. Attacks are in various forms as found in computer systems, networks and the cyber space. The immense increment in the amount of internet applications and the appearance of modern networks has created the need for improved security mechanisms. A good example of such modern technology is Internet of Things (IoTs). An IoT is a system that uses the Internet to facilitate communication between sensors and devices. Several approaches have been used to build attacks detection system in the past. The approaches for classifying attacks have been categorised as signature-based and Machine learning based. However, ML techniques have been argued to be more efficient for the identification of attacks or intrusions when compared to signature-based approaches. This study sourced for relevant literature from notable repositories and then surveyed some of the recent datasets that are very promising for ML-based studies in attack classification in IoT environments. The study equally provided a survey of evolving ML-based techniques for the classification of attacks in IoT networks. The study provided clear directions to researchers working in this area of researches by making the necessary information available more easily for the researcher to go about achieving improved ML-based approaches in this area.

With the rapid advancements and notable achievements across various application domains, Machine Learning (ML) has become a vital element within the Internet of Things (IoT) ecosystem. Among these use cases is IoT security, where numerous systems are deployed to identify or thwart attacks, including intrusion detection systems (IDSs), malware detection systems (MDSs), and device identification systems (DISs). Machine Learning-based (ML-based) IoT security systems can fulfill several security objectives, including detecting attacks, authenticating users before they gain access to the system, and categorizing suspicious activities. Nevertheless, ML faces numerous challenges, such as those resulting from the emergence of adversarial attacks crafted to mislead classifiers. This paper provides a comprehensive review of the body of knowledge about adversarial attacks and defense mechanisms, with a particular focus on three prominent IoT security systems: IDSs, MDSs, and DISs. The paper starts by establishing a taxonomy of adversarial attacks within the context of IoT. Then, various methodologies employed in the generation of adversarial attacks are described and classified within a two-dimensional framework. Additionally, we describe existing countermeasures for enhancing IoT security against adversarial attacks. Finally, we explore the most recent literature on the vulnerability of three ML-based IoT security systems to adversarial attacks.

Internet of Things (IoT) networks are being designed to connect billions of sensors/devices for an application-specific environment to provide services and carry out special tasks. It enables the virtual connection of the real-world physical devices with Internet cyberspace. While, there is an exponentially growing trend in the use of IoT technology for diverse applications such as smart agriculture, smart industries, smart hospitals, smart homes, etc., there are pressing design issues which affect the IoT network performance in the long run, for example, replacement of batteries for increasing the lifetime of devices to maintain the network longevity. To tackle this challenge, various wireless powered-based energy transfer paradigms were proposed over time. Backscatter communication (BackCom) is likewise one of the energy-efficient solutions proposed, which can satisfy the stringent green communication aspects of IoT networks. To achieve massive connectivity in BackCom-enabled IoT network, selecting a suitable multiple access scheme is also a crucial part. Most multiple access schemes proposed for BackCom systems are based on frequency or timeslots sharing techniques. Multiple access techniques proposed for BackCom IoT system can reap the gain of non-orthogonal multiple access (NOMA) principle. In NOMA, multiple end nodes (ENs) of sensors network is served in the same resource block, thereby, increasing the spectrum efficiency of BackCom IoT system. In this chapter, we discuss the prospects of backscatter technique with application of NOMA for green communication in IoT network.

Distributed Denial-of-Service (DDoS) attacks are widespread for Internet of Things (IoT) systems that aim to disrupt the availability of a system completely (high-rate DDoS) or partially (low-rate DDoS). Design and placement of intrusion detection systems (IDSs) for DDoS attacks on IoT systems are challenging due to the low power and lossy nature of networks. Existing IDSs are designed to handle either high-rate or low-rate DDoS but cannot handle both with good accuracy. Existing IDS placement techniques are mostly nontransparent, making malicious nodes aware of the presence of IDS nodes. Most of the IDS placement strategies are nonoptimal, making them energy inefficient. Accordingly, this work proposes a transparent, optimally placed, distributed IDS solution, namely, OPTIMIST, which can handle both high-rate and low-rate DDoS attacks with good accuracy. The placement problem is formulated as the weighted minimum vertex cover problem of a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${K}$ </tex-math></inline-formula> -uniform hypergraph and solved with an approximation algorithm. The IDS module is based on a long short-term memory (LSTM) model where a novel offline training method for LSTM is proposed using Wasserstein GAN-generated artificial flows. Extensive experimentation on simulation and testbed shows that the OPTIMIST can best achieve the balance between DDoS detection and energy overhead.

The environmental impacts that are associated with the interconnected sensor devices in the Internet of Things (IoT) network are minimized with the utilization of green communication. This is because, when the IoT devices tend to grow, the IoT system becomes more prone to issues regarding the utilization of resources, sustainability, and energy consumption within the communication system. Tuning the communication protocols, lowering the IoT device’s energy usage rate, and establishing an effective technique for the transmission of data are performed. This is achieved by the green communication technique, which aids in the establishment of significant energy savings in the IoT network. For devices with limited power sources, the green communication technology offers an extended battery life. Despite the enormous potential of IoT technology, a number of obstacles need to be overcome, including those related to load balancing, security, storage, privacy, energy management, and device heterogeneity. In response to the challenges posed by conventional models, an innovative solution for the selection of Cluster Head (CH) using a hybrid approach is devised here. The energy consumption of sensor nodes is influenced by various factors, such asCH load, temperature, distance, residual energy, the number of alive nodes, delay, and so on. To address these challenges in IoT networks, a hybrid approach that combines the Squid Game Optimizer (SGO) and the Artificial Gorilla Troops Optimizer (AGTO) for the optimal selection of CH is suggested. The implemented approach is named Hybrid Squid Game with Artificial Gorilla Troops Optimizer (HSG-AGTO). This approach aims to optimize the above-mentioned factors and overcome energy consumption challenges in IoT networks. The effectiveness of the model is validated, and the results demonstrate the superior performance of the suggested model.