International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 05 | May 2024
p-ISSN: 2395-0072
www.irjet.net
Satellite Communication Security: Evaluation of Anomaly Detection Models Clifa Mascarenhas1, Dr. Nilesh B. Fal Dessai2 1Student, Department of Information Technology and Engineering, Goa College of Engineering, Farmagudi, Goa,
India
2Head of Department, Department of Information Technology and Engineering, Goa College of Engineering,
Farmagudi, Goa, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Satellite-Based Communication (SATCOM)
Security Threat Landscape: While the project narrows its focus to anomaly detection, it remains cognizant of the broader security threat landscape. Cyber threats, deliberate interference, and signal jamming are acknowledged as potential sources of anomalies, prompting the need for advanced detection mechanisms.
systems are increasingly pivotal for connecting remote areas, driven by advancements in manufacturing and radio technologies. However, this surge in adoption comes with heightened cybersecurity challenges. Neglected security services, coupled with evolving cyber threats, create an expanded attack surface for SATCOM. The study systematically compares the performance of diverse anomaly detection models on a synthetic dataset, assessing their ability to identify and counteract deviations from normal system behavior.
In essence, this project serves as a focused exploration into the application of anomaly detection models to the security dimensions of SATCOM systems. By bridging theoretical analyses with practical considerations, this endeavor aims to contribute valuable insights to the ongoing discourse on securing and sustaining SATCOM systems in an ever-evolving digital landscape.
The outcomes of this research will contribute valuable insights to the selection of effective anomaly detection models for SATCOM security. Stakeholders involved in the SATCOM domain will benefit from this research, providing a foundation for informed decision-making when implementing security measures to safeguard critical communication channels.
2. Related Works The paper[1] employs a robust ground truth methodology, labeling a specific number of segments in each telemetry variable to establish the accuracy of anomaly detection methods. Out of 21,644 samples, 287 are labeled as fake anomalies. Notably, the Deviation Divide Mean over Neighbors (DDMN) consistently achieves 100% precision and maintains a high F1-score above 90%. DDMN's F1score reaches 0.969 when H = 2, surpassing DDM(next), DDM(prior), Z-SCORE, GMM, and K-means by significant margins of 93%, 92%, 30%, 61%, and 126%, respectively. Importantly, DDMN exhibits robustness to variations in the threshold (H). In contrast, other methods such as DDM(next), DDM(prior), and Z-SCORE exhibit low precision when H is small, as they tend to detect rapidly changing values as fake anomalies. GMM and K-means perform poorly in recall, erroneously assigning fake anomalies to the normal cluster.
Key Words: SATCOM, cybersecurity, anti-jamming, integrity, confidentiality, anomaly detection
1.INTRODUCTION In an era marked by the imperative of global connectivity, Satellite-Based Communication (SATCOM) systems have emerged as pivotal facilitators, connecting remote regions and fostering seamless communication. As the demand for robust connectivity continues to rise, this project embarks on a comprehensive study, focusing on the intricate landscape of security threats, solutions, and challenges inherent in the deployment and operation of SATCOM systems, with a specific emphasis on anomaly detection.
This paper proposes a data-driven anomaly detection framework for satellite telemetry with fake anomalies. The authors propose the Deviation Divide Mean over Neighbors (DDMN) method to solve the fake anomaly problem caused by data errors in satellite telemetry data. They then use Long Short-Term Memory (LSTM), a deep learning method, to model multivariable time-series data and a Gaussian model to detect anomalies.
Anomaly Detection Models: The primary objective of this project is to explore and evaluate various anomaly detection models tailored for SATCOM systems. Anomalies, deviations from the expected behavior, can signify potential security breaches or operational irregularities. This study delves into the effectiveness of state-of-the-art models, assessing their precision, recall, and overall performance in identifying anomalies within SATCOM data.
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 626