Intrusion Detection in Software-Defined Networks: Leveraging Deep Reinforcement Learning with Graph Convolutional Networks for Resilient Infrastructure
Fuqdan A. Al-Ibraheemi1 , Firas Hazzaa2,3, Mohanad Sameer Jabbar 4,*, Jamal Fadhil Tawfeq 5, Ravi Sekhar 6, Pritesh Shah 7, Sushma Parihar 8
1College of Dentistry, University of Al-Ameed, Iraq
2 Ministry of Higher Education and Scientific Research , Baghdad, Iraq
3 Visiting Fellow , School of Engineering and Build Environment, Anglia Ruskin University, Chelmsford , UK
4Medical Instruments techniques Engineering Department, Technical College of Engineering, Al-Bayan University, Baghdad, Iraq
5 Department of Medical Instrumentation Technical Engineering, Medical Technical College, Al-Farahidi University, Baghdad, Iraq
6,7,8 Symbiosis Institute of Technology (SIT) Pune Campus, Symbiosis International (Deemed University) (SIU), Pune, 412115, Maharashtra, India
Emails: fuqdanal_ibrahimi@alameed.edu.iq; fih7600@gmail.com; mohanad.s@albayan.edu.iq; jamaltawfeq55@gmail.com; ravi.sekhar@sitpune.edu.in; pritesh.shah@sitpune.edu.in; sushmap@sitpune.edu.in
Abstract
In the realm of academic technology, a Virtual Learning Environment (VLE) serves as a web-based platform designed to facilitate the digital aspects of educational curricula, primarily within educational institutions. It encompasses various stages of assessment and provides resources, assignments, and interactive elements within a structured course framework. Its utilization became particularly prominent during the pandemic, as it proved highly beneficial to students by delivering cost-effective and flexible remote learning options. However, despite its advantages, VLEs come with notable limitations. Human emotion and awareness must be evaluated in virtual learning environments to improve user experience. A Fuzzy-based Convolutional Neural Network (FCNN) has been proposed to identify human emotions in a virtual learning environment. Our evaluation of the virtual learning environment's awareness is based on data collected through questionnaire surveys. Face images are preprocessed using histogram equalization. DCT allows a high-level feature extraction process. In addition, AFCNNs allow virtual learners to assess emotions and awareness efficiently. Using this approach, we evaluate accuracy, sensitivity, specificity, and precision. By comparing our proposed educational system's performance to those of traditional sustainable development education, we prove the effectiveness of our proposal.
Keywords: Computer Science; Network; Virtual Learning Environment (VLE); Fuzzy-based Convolutional Neural Network (FCNN); Software-Defined Networking (SDN).