Biblio
Supervisory control and data acquisition (SCADA) networks provide high situational awareness and automation control for industrial control systems, whilst introducing a wide range of access points for cyber attackers. To address these issues, a line of machine learning or deep learning based intrusion detection systems (IDSs) have been presented in the literature, where a large number of attack examples are usually demanded. However, in real-world SCADA networks, attack examples are not always sufficient, having only a few shots in many cases. In this paper, we propose a novel few-shot learning based IDS, named FS-IDS, to detect cyber attacks against SCADA networks, especially when having only a few attack examples in the defenders’ hands. Specifically, a new method by orchestrating one-hot encoding and principal component analysis is developed, to preprocess SCADA datasets containing sufficient examples for frequent cyber attacks. Then, a few-shot learning based preliminary IDS model is designed and trained using the preprocessed data. Last, a complete FS-IDS model for SCADA networks is established by further training the preliminary IDS model with a few examples for cyber attacks of interest. The high effectiveness of the proposed FS-IDS, in detecting cyber attacks against SCADA networks with only a few examples, is demonstrated by extensive experiments on a real SCADA dataset.
In recent times, an increasing amount of intelligent electronic devices (IEDs) are being deployed to make power systems more reliable and economical. While these technologies are necessary for realizing a cyber-physical infrastructure for future smart power grids, they also introduce new vulnerabilities in the grid to different cyber-attacks. Traditional methods such as state vector estimation (SVE) are not capable of identifying cyber-attacks while the geometric information is also injected as an attack vector. In this paper, a machine learning based smart grid attack identification method is proposed. The proposed method is carried out by first collecting smart grid power flow data for machine learning training purposes which is later used to classify the attacks. The performance of both the proposed SVM method and the traditional SVE method are validated on IEEE 14, 30, 39, 57 and 118 bus systems, and the performance regarding the scale of the power system is evaluated. The results show that the SVM-based method performs better than the SVE-based in attack identification over a much wider scale of power systems.
Cyber attacks and the associated costs made cybersecurity a vital part of any system. User behavior and decisions are still a major part in the coping with these risks. We developed a model of optimal investment and human decisions with security measures, given that the effectiveness of each measure depends partly on the performance of the others. In an online experiment, participants classified events as malicious or non-malicious, based on the value of an observed variable. Prior to making the decisions, they had invested in three security measures - a firewall, an IDS or insurance. In three experimental conditions, maximal investment in only one of the measures was optimal, while in a fourth condition, participants should not have invested in any of the measures. A previous paper presents the analysis of the investment decisions. This paper reports users' classifications of events when interacting with these systems. The use of security mechanisms helped participants gain higher scores. Participants benefited in particular from purchasing IDS and/or Cyber Insurance. Participants also showed higher sensitivity and compliance with the alerting system when they could benefit from investing in the IDS. Participants, however, did not adjust their behavior optimally to the security settings they had chosen. The results demonstrate the complex nature of risk-related behaviors and the need to consider human abilities and biases when designing cyber security systems.
Machine learning techniques help to understand underlying patterns in datasets to develop defense mechanisms against cyber attacks. Multilayer Perceptron (MLP) technique is a machine learning technique used in detecting attack vs. benign data. However, it is difficult to construct any effective model when there are imbalances in the dataset that prevent proper classification of attack samples in data. In this research, we use UGR'16 dataset to conduct data wrangling initially. This technique helps to prepare a test set from the original dataset to train the neural network model effectively. We experimented with a series of inputs of varying sizes (i.e. 10000, 50000, 1 million) to observe the performance of the MLP neural network model with distribution of features over accuracy. Later, we use Generative Adversarial Network (GAN) model that produces samples of different attack labels (e.g. blacklist, anomaly spam, ssh scan) for balancing the dataset. These samples are generated based on data from the UGR'16 dataset. Further experiments with MLP neural network model shows that a balanced attack sample dataset, made possible with GAN, produces more accurate results than an imbalanced one.
Modern industrial control systems (ICS) act as victims of cyber attacks more often in last years. These cyber attacks often can not be detected by classical information security methods. Moreover, the consequences of cyber attack's impact can be catastrophic. Since cyber attacks leads to appearance of anomalies in the ICS and technological equipment controlled by it, the task of intrusion detection for ICS can be reformulated as the task of industrial process anomaly detection. This paper considers the applicability of generative adversarial networks (GANs) in the field of industrial processes anomaly detection. Existing approaches for GANs usage in the field of information security (such as anomaly detection in network traffic) were described. It is proposed to use the BiGAN architecture in order to detect anomalies in the industrial processes. The proposed approach has been tested on Secure Water Treatment Dataset (SWaT). The obtained results indicate the prospects of using the examined method in practice.
Wide integration of information and communication technology (ICT) in modern power grids has brought many benefits as well as the risk of cyber attacks. A critical step towards defending grid cyber security is to understand the cyber-physical causal chain, which describes the progression of intrusion in cyber-space leading to the formation of consequences on the physical power grid. In this paper, we develop an attack vector for a time delay attack at load frequency control in the power grid. Distinct from existing works, which are separately focused on cyber intrusion, grid response, or testbed validation, the proposed attack vector for the first time provides a full cyber-physical causal chain. It targets specific vulnerabilities in the protocols, performs a denial-of-service (DoS) attack, induces the delays in control loop, and destabilizes grid frequency. The proposed attack vector is proved in theory, presented as an attack tree, and validated in an experimental environment. The results will provide valuable insights to develop security measures and robust controls against time delay attacks.
The increase of cyber attacks in both the numbers and varieties in recent years demands to build a more sophisticated network intrusion detection system (NIDS). These NIDS perform better when they can monitor all the traffic traversing through the network like when being deployed on a Software-Defined Network (SDN). Because of the inability to detect zero-day attacks, signature-based NIDS which were traditionally used for detecting malicious traffic are beginning to get replaced by anomaly-based NIDS built on neural networks. However, recently it has been shown that such NIDS have their own drawback namely being vulnerable to the adversarial example attack. Moreover, they were mostly evaluated on the old datasets which don't represent the variety of attacks network systems might face these days. In this paper, we present Reconstruction from Partial Observation (RePO) as a new mechanism to build an NIDS with the help of denoising autoencoders capable of detecting different types of network attacks in a low false alert setting with an enhanced robustness against adversarial example attack. Our evaluation conducted on a dataset with a variety of network attacks shows denoising autoencoders can improve detection of malicious traffic by up to 29% in a normal setting and by up to 45% in an adversarial setting compared to other recently proposed anomaly detectors.
In the process of informationization and networking of smart grids, the original physical isolation was broken, potential risks increased, and the increasingly serious cyber security situation was faced. Therefore, it is critical to develop accuracy and efficient anomaly detection methods to disclose various threats. However, in the industry, mainstream security devices such as firewalls are not able to detect and resist some advanced behavior attacks. In this paper, we propose a time series anomaly detection model, which is based on the periodic extraction method of discrete Fourier transform, and determines the sequence position of each element in the period by periodic overlapping mapping, thereby accurately describe the timing relationship between each network message. The experiments demonstrate that our model can detect cyber attacks such as man-in-the-middle, malicious injection, and Dos in a highly periodic network.
The limited information on the cyberattacks available in the unclassified regime, hardens standardizing the analysis. We address the problem of modeling and analyzing cyberattacks using a multimodal graph approach. We formulate the stages, actors, and outcomes of cyberattacks as a multimodal graph. Multimodal graph nodes include cyberattack victims, adversaries, autonomous systems, and the observed cyber events. In multimodal graphs, single-modality graphs are interconnected according to their interaction. We apply community and centrality analysis on the graph to obtain in-depth insights into the attack. In community analysis, we cluster those nodes that exhibit “strong” inter-modal ties. We further use centrality to rank the nodes according to their importance. Classifying nodes according to centrality provides the progression of the attack from the attacker to the targeted nodes. We apply our methods to two popular case studies, namely GhostNet and Putter Panda and demonstrate a clear distinction in the attack stages.