With the rapid development and popularization of Internet of Things (IoT) devices, an increasing number of cyber-attacks are targeting such devices. It was said that most of the attacks in IoT environments are botnet-based attacks. Many security weaknesses still exist on the IoT devices because most of them have not enough memory and computational resource for robust security mechanisms. Moreover, many existing rule-based detection systems can be circumvented by attackers. In this study, we proposed a machine learning (ML)-based botnet attack detection framework with sequential detection architecture. An efficient feature selection approach is adopted to implement a lightweight detection system with a high performance. The overall detection performance achieves around 99% for the botnet attack detection using three different ML algorithms, including artificial neural network (ANN), J48 decision tree, and Naïve Bayes. The experiment result indicates that the proposed architecture can effectively detect botnet-based attacks, and also can be extended with corresponding sub-engines for new kinds of attacks.

The application of a large number of Internet of Things (IoT) devices makes our life more convenient and industries more efficient. However, it also makes cyber-attacks much easier to occur because so many IoT devices are deployed and most of them do not have enough resources (i.e., computation and storage capacity) to carry out ordinary intrusion detection systems (IDSs). In this study, a lightweight machine learning-based IDS using a new feature selection algorithm is designed and implemented on Raspberry Pi, and its performance is verified using a public dataset collected from an IoT environment. To make the system lightweight, we propose a new algorithm for feature selection, called the correlated-set thresholding on gain-ratio (CST-GR) algorithm, to select really necessary features. Because the feature selection is conducted on three specific kinds of cyber-attacks, the number of selected features can be significantly reduced, which makes the classifiers very small and fast. Thus, our detection system is lightweight enough to be implemented and carried out in a Raspberry Pi system. More importantly, as the really necessary features corresponding to each kind of attack are exploited, good detection performance can be expected. The performance of our proposal is examined in detail with different machine learning algorithms, in order to learn which of them is the best option for our system. The experiment results indicate that the new feature selection algorithm can select only very few features for each kind of attack. Thus, the detection system is lightweight enough to be implemented in the Raspberry Pi environment with almost no sacrifice on detection performance.

Cyber attack detection technology plays a vital role today, since cyber attacks have been causing great harm and loss to organizations and individuals. Feature selection is a necessary step for many cyber-attack detection systems, because it can reduce training costs, improve detection performance, and make the detection system lightweight. Many techniques related to feature selection for cyber attack detection have been proposed, and each technique has advantages and disadvantages. Determining which technology should be selected is a challenging problem for many researchers and system developers, and although there have been several survey papers on feature selection techniques in the field of cyber security, most of them try to be all-encompassing and are too general, making it difficult for readers to grasp the concrete and comprehensive image of the methods. In this paper, we survey the filter-based feature selection technique in detail and comprehensively for the first time. The filter-based technique is one popular kind of feature selection technique and is widely used in both research and application. In addition to general descriptions of this kind of method, we also explain in detail search algorithms and relevance measures, which are two necessary technical elements commonly used in the filter-based technique.

Sensor-based Human Activity Recognition (HAR) is crucial in ubiquitous computing, analyzing behaviors through multi-dimensional observations. Despite research progress, HAR confronts challenges, particularly in data distribution assumptions. Most studies assume uniform data distributions across datasets, contrasting with the varied nature of practical sensor data in human activities. Addressing data heterogeneity issues can improve performance, reduce computational costs, and aid in developing personalized, adaptive models with fewer annotated data. This review investigates how machine learning addresses data heterogeneity in HAR by categorizing data heterogeneity types, applying corresponding suitable machine learning methods, summarizing available datasets, and discussing future challenges.

A current trend of research focuses on artificial intelligence based cryptography which although proposed almost thirty years ago could not attract much attention. Abadi and Anderson’s work on adversarial cryptography in 2016 rejuvenated the research area which now focuses in building neural networks that are able to learn cryptography using the idea from Generative Adversarial Networks (GANs). In this paper, we survey the most prominent research works that cover neural networks based cryptography from two main periods. The first period covers the oldest models that have been proposed shortly after 2000 and the second period covers the more recent models that have been proposed since 2016. We first discuss the implementation of the systems from the earlier era and the attacks mounted on them. After that, we focus on post 2016 era where more advanced techniques are utilized that rely on GANs in which neural networks compete with each other in order to achieve a goal e.g. learning to encrypt a communication. Finally, we discuss security analysis performed on adversarial cryptography models.