AI-Driven Cybersecurity for Defense Networks: A Mathematical Approach with DDoS Attack Analysis

Neelesh, Mungoli (2025) AI-Driven Cybersecurity for Defense Networks: A Mathematical Approach with DDoS Attack Analysis. International Journal of Trend in Scientific Research and Development, 9 (2). pp. 206-223. ISSN 2456-6470

Text ijtsrd76306.pdf Download (2MB)

Abstract

This paper presents a novel machine-learning pipeline tailored for proactive cyber defense within high-stakes military networks, addressing the pressing need to detect and neutralize sophisticated threats such as Distributed Denial-of-Service (DDoS) attacks in near real-time. Our approach begins with a mathematically rigorous anomaly detection framework, constructed on the premise that normal network traffic follows an identifiable statistical distribution, deviations from which can serve as early indicators of malicious behavior. By exploiting deep neural architectures—specifically autoencoders enhanced with domain-specific heuristics—our pipeline learns complex traffic patterns, encompassing both high-volume and subtle “low-and-slow” attack methodologies. A core component of our methodology involves deriving explicit theoretical bounds for detection accuracy and false alarm rates, ensuring that defense operators can calibrate the system according to mission-critical thresholds.

Item Type:	Article
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	Postgraduate > Master's of Islamic Education
Depositing User:	Journal Editor
Date Deposited:	15 Mar 2025 11:51
Last Modified:	15 Mar 2025 11:51
URI:	http://eprints.umsida.ac.id/id/eprint/15832

Actions (login required)

View Item