More Like this

1. Network Intrusion Detection in Smart Grids for Imbalanced Attack Types Using Machine Learning Models

   https://doi.org/10.1109/ICTC46691.2019.8939744  

   Roy, Dipanjan Das; Shin, Dongwan (October 2019, 2019 International Conference on Information and Communication Technology Convergence (ICTC))

   Smart grid has evolved as the next generation power grid paradigm which enables the transfer of real time information between the utility company and the consumer via smart meter and advanced metering infrastructure (AMI). These information facilitate many services for both, such as automatic meter reading, demand side management, and time-of-use (TOU) pricing. However, there have been growing security and privacy concerns over smart grid systems, which are built with both smart and legacy information and operational technologies. Intrusion detection is a critical security service for smart grid systems, alerting the system operator for the presence of ongoing attacks. Hence, there has been lots of research conducted on intrusion detection in the past, especially anomaly-based intrusion detection. Problems emerge when common approaches of pattern recognition are used for imbalanced data which represent much more data instances belonging to normal behaviors than to attack ones, and these approaches cause low detection rates for minority classes. In this paper, we study various machine learning models to overcome this drawback by using CIC-IDS2018 dataset [1]. 

   more » « less
2. Virtual Machine Introspection for Anomaly-Based Keylogger Detection

   https://doi.org/10.1109/HPSR48589.2020.9098980  

   Huseynov, H.; Kouraiy, H.; Saadawi, T.; Igbe, O. (May 2020, IEEE High Performance Switching and Routing)

   Software Keyloggers are dominant class of malicious applications that surreptitiously logs all the user activity to gather confidential information. Among many other types of keyloggers, API-based keyloggers can pretend as unprivileged program running in a user-space to eavesdrop and record all the keystrokes typed by the user. In a Linux environment, defending against these types of malware means defending the kernel against being compromised and it is still an open and difficult problem. Considering how recent trend of edge computing extends cloud computing and the Internet of Things (IoT) to the edge of the network, a new types of intrusiondetection system (IDS) has been used to mitigate cybersecurity threats in edge computing. Proposed work aims to provide secure environment by constantly checking virtual machines for the presence of keyloggers using cutting edge artificial immune system (AIS) based technology. The algorithms that exist in the field of AIS exploit the immune system’s characteristics of learning and memory to solve diverse problems. We further present our approach by employing an architecture where host OS and a virtual machine (VM) layer actively collaborate to guarantee kernel integrity. This collaborative approach allows us to introspect VM by tracking events (interrupts, system calls, memory writes, network activities, etc.) and to detect anomalies by employing negative selection algorithm (NSA). 

   more » « less
3. Adversarial Anomaly Detection Using Centroid-Based Clustering

   https://doi.org/10.1109/IRI.2018.00009  

   Anindya, Imrul Chowdhury; Kantarcioglu, Murat (July 2018, 2018 IEEE International Conference on Information Reuse and Integration (IRI))

   As cyber attacks are growing with an unprecedented rate in the recent years, organizations are seeking an efficient and scalable solution towards a holistic protection system. As the adversaries are becoming more skilled and organized, traditional rule based detection systems have been proved to be quite ineffective against the continuously evolving cyber attacks. Consequently, security researchers are focusing on applying machine learning techniques and big data analytics to defend against cyber attacks. Over the recent years, several anomaly detection systems have been claimed to be quite successful against the sophisticated cyber attacks including the previously unseen zero-day attacks. But often, these systems do not consider the adversary's adaptive attacking behavior for bypassing the detection procedure. As a result, deploying these systems in active real-world scenarios fails to provide significant benefits in the presence of intelligent adversaries that are carefully manipulating the attack vectors. In this work, we analyze the adversarial impact on anomaly detection models that are built upon centroid-based clustering from game-theoretic aspect and propose adversarial anomaly detection technique for these models. The experimental results show that our game-theoretic anomaly detection models can withstand attacks more effectively compared to the traditional models. 

   more » « less
4. Detecting Cyber Attacks in Smart Grids Using Semi-Supervised Anomaly Detection and Deep Representation Learning

   https://doi.org/10.3390/info12080328  

   Qi, Ruobin; Rasband, Craig; Zheng, Jun; Longoria, Raul (August 2021, Information)

   null (Ed.)

   Smart grids integrate advanced information and communication technologies (ICTs) into traditional power grids for more efficient and resilient power delivery and management, but also introduce new security vulnerabilities that can be exploited by adversaries to launch cyber attacks, causing severe consequences such as massive blackout and infrastructure damages. Existing machine learning-based methods for detecting cyber attacks in smart grids are mostly based on supervised learning, which need the instances of both normal and attack events for training. In addition, supervised learning requires that the training dataset includes representative instances of various types of attack events to train a good model, which is sometimes hard if not impossible. This paper presents a new method for detecting cyber attacks in smart grids using PMU data, which is based on semi-supervised anomaly detection and deep representation learning. Semi-supervised anomaly detection only employs the instances of normal events to train detection models, making it suitable for finding unknown attack events. A number of popular semi-supervised anomaly detection algorithms were investigated in our study using publicly available power system cyber attack datasets to identify the best-performing ones. The performance comparison with popular supervised algorithms demonstrates that semi-supervised algorithms are more capable of finding attack events than supervised algorithms. Our results also show that the performance of semi-supervised anomaly detection algorithms can be further improved by augmenting with deep representation learning. 

   more » « less
5. Pluto: Sample Selection for Robust Anomaly Detection on Polluted Log Data

   Ma, Lei; Cao, Lei; VanNostrand, Peter M; Hofmann, Dennis M; Su, Yao; Rundensteiner, Elke A (September 2024, Proceedings of the ACM on Management of Data)

   Log anomaly detection, critical in identifying system failures and preempting security breaches, finds irregular patterns within large volumes of log data. Modern log anomaly detectors rely on training deep learning models on clean anomaly-free log data. However, such clean log data requires expensive and tedious human labeling. In this paper, we thus propose a robust log anomaly detection framework, PlutoNOSPACE, that automatically selects a clean representative sample subset of the polluted log sequence data to train a Transformer-based anomaly detection model. Pluto features three innovations. First, due to localized concentrations of anomalies inherent in the embedding space of log data, Pluto partitions the sequence embedding space generated by the model into regions that then allow it to identify and discard regions that are highly polluted by our pollution level estimation scheme, based on our pollution quantification via Gaussian mixture modeling. Second, for the remaining more slightly polluted regions, we select samples that maximally purify the eigenvector spectrum, which can be transformed into the NP-hard facility location problem; allowing us to leverage its greedy solution with a (1-(1/e)) approximation guarantee in optimality. Third, by iteratively alternating between the above subset selection, a model re-training on the latest subset, and a subset filtering using dynamic training artifacts generated by the latest model, the data selected is progressively refined. The final sample set is used to retrain the final anomaly detection model. Our experiments on four real-world log benchmark datasets demonstrate that by retaining 77.7\% (BGL) to 96.6\% (ThunderBird) of the normal sequences while effectively removing 90.3\% (BGL) to 100.0\% (ThunderBird, HDFS) of the anomalies, Pluto provides a significant absolute F-1 improvement up to 68.86\% (2.16\% → 71.02\%) compared to the state-of-the-art sample selection methods. The implementation of this work is available at https://github.com/LeiMa0324/Pluto-SIGMOD25. 

   more » « less