Article
Computer Science, Information Systems
Wai Weng Lo, Gayan Kulatilleke, Mohanad Sarhan, Siamak Layeghy, Marius Portmann
Summary: In this paper, we propose XG-BoT, an explainable deep graph neural network model for botnet node detection. The model comprises a botnet detector and an explainer for automatic forensics. XG-BoT detector effectively detects malicious botnet nodes, while the explainer highlights suspicious network flows and botnet nodes for automatic network forensics.
INTERNET OF THINGS
(2023)
Article
Computer Science, Information Systems
Islam Debicha, Benjamin Cochez, Tayeb Kenaza, Thibault Debatty, Jean -Michel Dricot, Wim Mees
Summary: Due to the vulnerability of machine learning algorithms, many studies have shown that they can be fooled by adversarial attacks. This study aims to investigate the feasibility of such attacks against network-based intrusion detection systems and propose a defensive scheme to protect ML-based IDSs.
COMPUTERS & SECURITY
(2023)
Article
Computer Science, Hardware & Architecture
Vicente Quezada, Fabian Astudillo-Salinas, Luis Tello-Oquendo, Paul Bernal
Summary: In today's cyberattacks, botnets are used for sophisticated and coordinated attacks. This research presents a bot infection detection system based on DNS traffic events. Using fingerprint analysis and Isolation Forest, infected hosts can be identified. The system also utilizes Random Forest to detect future bot infections with a high precision of over 99%.
Article
Telecommunications
Chirag Joshi, Ranjeet K. Ranjan, Vishal Bharti
Summary: With the growth of Internet and wireless communication, Internet-of-Things (IoT) has become a significant technology for smart devices, but the always-on connectivity and complexity of distributed computing have made IoT systems poorly secured and vulnerable to malicious attacks, including botnets. This paper proposes a novel feature selection approach based on Ant Colony optimization algorithm and Artificial Neural Network for IoT-Botnet detection, achieving high accuracy of 99.68% and improving accuracy by 5% through feature selection.
WIRELESS PERSONAL COMMUNICATIONS
(2023)
Article
Computer Science, Information Systems
Dandy Pramana Hostiadi, Tohari Ahmad
Summary: This paper proposes a new model for detecting bot group activity using a hybrid analysis approach, including extracting activity patterns, analyzing activity similarities, and analyzing activity correlation. The experiment results show that the proposed method can detect bot group activity with a high accuracy of 99.73% and a false-positive rate of less than 1%.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Khlood Shinan, Khalid Alsubhi, M. Usman Ashraf
Summary: This study aims to build a graph-based botnet malware detection system utilizing centrality measures and machine learning. By applying efficient centrality measures, the proposed system is robust, achieving 99% accuracy in detecting botnet attacks with a false positive rate as low as 0.0001%.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Information Systems
Abdelouahid Derhab, Rahaf Alawwad, Khawlah Dehwah, Noshina Tariq, Farrukh Aslam Khan, Jalal Al-Muhtadi
Summary: Twitter, a popular micro-blogging platform, faces risks such as spreading rumors and malware. Tweet-based botnets pose a threat to users, necessitating the use of big data analytics and shallow/deep learning techniques to accurately distinguish between human and bot accounts.
Article
Computer Science, Information Systems
Qinglin He, Lihong Wang, Lin Cui, Libin Yang, Bing Luo
Summary: This paper proposes a gravity-based critical bots identification scheme that can assess and identify important bots in large-scale botnet infections through modeling and algorithm design.
Article
Computer Science, Information Systems
Sajad Hamzenejadi, Mahdieh Ghazvini, Seyedamiryousef Hosseini
Summary: With the increasing number of people using mobile devices, mobile devices have become prime targets for cybercriminals. This paper provides a detailed background on mobile botnets and discusses various techniques for detecting them.
INTERNATIONAL JOURNAL OF INFORMATION SECURITY
(2023)
Article
Computer Science, Artificial Intelligence
Grazyna Suchacka, Alberto Cabri, Stefano Rovetta, Francesco Masulli
Summary: This study presents a novel method for real-time identification of bots, using machine learning and probabilistic analysis to quickly classify active sessions as "bots" or "humans". Empirical research confirms the effectiveness of this approach in distinguishing between bots and human visitors at early stages.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Hardware & Architecture
Tong Anh Tuan, Nguyen Viet Anh, Tran Thi Luong, Hoang Viet Long
Summary: This study introduces a new dataset on DGA botnets named UTL_DGA22, which addresses the detection and classification problems in cybersecurity. The dataset includes only domain records and proposed a valuable set of attributes for classification algorithms, leading to good results in experiments. The UTL_DGA22 dataset serves as a database for researchers to develop algorithms and evaluate solutions objectively.
Article
Computer Science, Theory & Methods
Efe Arin, Mucahid Kutlu
Summary: While social bots can have positive impacts, they can also be used for malicious purposes. Detecting bots on social media platforms is crucial, especially with the development of AI making bots more sophisticated. This study proposes a deep learning architecture with LSTM models and a fully connected layer to analyze social media activity and successfully outperforms other baselines in experiments.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2023)
Article
Chemistry, Multidisciplinary
Shema Alosaimi, Saad M. Almutairi
Summary: The rapid growth of IoT has improved the quality of our lives through increased automation and interconnectivity, but the security of IoT devices remains a significant concern. This study proposes a novel approach combining deep learning and three-level algorithms to quickly and accurately detect attacks in IoT networks. Evaluation using the Bot-IoT dataset shows significant improvements in detection performance compared to existing methods. The proposed approach also has the potential to enhance the security of other IoT applications, making it a promising contribution to IoT security.
APPLIED SCIENCES-BASEL
(2023)
Article
Multidisciplinary Sciences
Ahmed Bahaa, Abdalla Sayed, Laila Elfangary, Hanan Fahmy
Summary: This study proposes a novel hybrid meta-heuristic adaptive optimization algorithm for optimizing the hyperparameters of a convolutional neural network, and compares it with other optimization algorithms. The results show that the proposed algorithm performs well in detecting IoT network attacks.
Proceedings Paper
Computer Science, Theory & Methods
P. Sai Teja, P. Hema Sirija, P. Roshini, S. Saravanan
Summary: The botnet, which can be client-server architecture or peer-to-peer architecture, is a group of malware-infected devices that perform malicious activities over the Internet. This paper proposes a Hadoop-based P2P botnet detection system that can effectively detect P2P bots in a local area network without converting PCAP files to text. Detection is based on various characteristics of P2P bots.
APPLIED SOFT COMPUTING AND COMMUNICATION NETWORKS
(2021)