Securing data center with digital twins

Foo, Kar Yeng (2025) Securing data center with digital twins. Final Year Project, UTAR.

PDF Download (3104Kb)

Abstract

This development-based project proposes an innovative approach to data center network planning by implementing Digital Twin technology to automate and accelerate network topology generation. Traditional network modelling tools, such as GNS3, rely on manual drag-and-drop configurations, which are inefficient and time-consuming for large-scale deployments. To address this limitation, this project introduces an AI-driven network diagram generator that leverages the OpenAI API to interpret natural language commands and dynamically construct network diagrams. Users can intuitively create a complete network, add, modify, or remove network components through text-based inputs, significantly reducing design time and minimizing human error. The system supports an export function that allows users to save generated network diagrams in both .png format for documentation and .json format for direct integration with GNS3. This enables seamless transfer of AI-generated topologies into a real emulation environment, allowing engineers to validate and refine their designs without rebuilding them manually. Compared to conventional methods, the proposed system provides a much faster and more efficient way to generate complex network topologies. The project follows a structured development methodology, including requirements analysis, prototype design, API integration, and iterative testing. Expected outcomes include a user-friendly, scalable tool that enhances network planning speed, efficiency, and compatibility with industry-standard platforms. By bridging the gap between AI automation and network emulation tools, this solution provides IT professionals with a smarter and more productive approach to data center infrastructure modelling.

Actions (login required)