Development of NetBot-AI: A Holistic Tool for Detecting and Mitigating VoIP Network Security Threats Using Hybrid Datasets
Year: 2025
Type: article
Abstract: The rapid adoption of Voice over IP (VoIP) has brought cost-effective, scalable communication, but also introduce security concerns, especially in areas with legacy infrastructure. Traditional methods struggle with emerging threats like SIP flooding and toll fraud. This study introduces NetBot-AI, an AI-powered system for real-time threat detection and adaptive response in VoIP networks. Using hybrid datasets and advanced pre-processing techniques, it employs Deep Q-Networks, ML classifiers, and the Ekolama (Eₖ) loss function to boost accuracy. Evaluated on 10 million call records, it achieved 98% binary and 95% multi-class accuracy with minimal false positives, reducing downtime by 40%. Its modular design ensures easy integration. Results show NetBot-AI outperforms traditional tools, offering scalable, proactive security for diverse telecom environments.
Authors: Bodunrin Isa Bakare, Solomon Malcolm Ekolama, Sunny Orike, Christopher Ahiakwo
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7082381398
FWCI:
Cites:
Cited by:
Related to: