Article: Networks intrusion detection system using particle reef optimisation-based convolutional Elman neural network Journal: International Journal of Networking and Virtual Organisations (IJNVO) 2024 Vol.31 No.4 pp.339 - 367 Abstract: This research introduces a novel strategy that harnesses the capabilities of two optimisation models, namely particle swarm optimisation (PSO) and coral reef optimisation (CRO). This approach is integrated with a hybrid framework that unites convolutional neural networks (CNN). Furthermore, the fusion of CNNs and Elman neural networks (ENNs) within the hybrid neural network architecture capitalises on their complementary strengths. CNNs excel at spatial feature extraction, effectively capturing patterns in network data, while ENNs specialise in modelling temporal dependencies within sequences of network events. This integration guarantees a thorough examination of network traffic, empowering the identification of both real-time and historically contextualised security threats. Practical findings substantiate the effectiveness of the particle reef optimisation-based convolutional Elman neural network (PRO-CENN) method. The achievement of the proposed system shows an accuracy of 97.27%, sensitivity of 92.87%, and specificity of 97.90% for 90% training percent and 96.28% accuracy, 95.64% sensitivity, and 96.92% specificity for 10 k-fold. Inderscience Publishers - linking academia, business and industry through research

Title: Networks intrusion detection system using particle reef optimisation-based convolutional Elman neural network

Authors: Kshyamasagar Mahanta; Hima Bindu Maringanti

Addresses: Department of Computer Application, Maharaja Sriram Chandra Bhanja Deo University, Takatpur, Baripada, Odisha, 757003, India ' Department of Computer Application, Maharaja Sriram Chandra Bhanja Deo University, Takatpur, Baripada, Odisha, 757003, India

Abstract: This research introduces a novel strategy that harnesses the capabilities of two optimisation models, namely particle swarm optimisation (PSO) and coral reef optimisation (CRO). This approach is integrated with a hybrid framework that unites convolutional neural networks (CNN). Furthermore, the fusion of CNNs and Elman neural networks (ENNs) within the hybrid neural network architecture capitalises on their complementary strengths. CNNs excel at spatial feature extraction, effectively capturing patterns in network data, while ENNs specialise in modelling temporal dependencies within sequences of network events. This integration guarantees a thorough examination of network traffic, empowering the identification of both real-time and historically contextualised security threats. Practical findings substantiate the effectiveness of the particle reef optimisation-based convolutional Elman neural network (PRO-CENN) method. The achievement of the proposed system shows an accuracy of 97.27%, sensitivity of 92.87%, and specificity of 97.90% for 90% training percent and 96.28% accuracy, 95.64% sensitivity, and 96.92% specificity for 10 k-fold.

Keywords: IDSs; intrusion detection systems; network security; CNN; convolutional neural networks; ENN; Elman neural networks; particle reef optimisation.

DOI: 10.1504/IJNVO.2024.144078

International Journal of Networking and Virtual Organisations, 2024 Vol.31 No.4, pp.339 - 367

Received: 04 May 2024
Accepted: 27 Aug 2024
Published online: 24 Jan 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Networks intrusion detection system using particle reef optimisation-based convolutional Elman neural network

Keep up-to-date