Malicious attacks detection using GRU-BWFA classifier in pervasive computing. (January 2023)

• Record Type: Journal Article
• Authors: Rajasekaran, P.; Magudeeswaran, V.
• Is Part Of: Biomedical signal processing and control. Volume 79(2023)Part 2
• Publication Date: 2023-01
• Subjects: Pervasive Computing -- DDoS attacks -- UNSW-NB15 dataset and NSL-KDD dataset -- Gated Recurrent Unit-Neural Networks based on Bidirectional weighted average classifier -- Enhanced Salp Swarm Optimization (ESSO); Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104219
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: DDoS attack detection is significant to secure a network or system for which the study uses machine learning algorithms. Enhanced Salp Swarm Optimization is proposed for feature selection to select only the optimal features. Gated Recurrent Unit Neural Network based on Bidirectional Weighted Feature Averaging (GRU-BWFA) classifier is employed for high detection rate in distinguishing the DDoS attacks. Abstract: With the advanced trends in pervasive computing, the data users were subjected to face different kinds of attacks. Several algorithms for attack detection was performed with minimal accuracy in prediction and consideration of performance metrics was not effective. Hence effective and prompt detection of malicious attacks must be optimized in terms of confidentiality, privacy, availability and integrity. Accordingly, the proposed research paper provides an effective mechanism for detecting and classifying DDoS attacks such as TCP-SYN, UDP flood, ICMP echo, HTTP flood, Slow Loris Slow Post and Brute Force attack, by utilizing machine learning methods within the UNSW-NB15 dataset and NSL-KDD dataset. Significantly, Gated Recurrent Unit Neural Network based on Bidirectional Weighted Feature Averaging (GRU-BWFA) classifier is utilized as a proposed classifier approach for high detection rate and accuracy in distinguishing the mentioned DDoS attacks. Feature selection is performed using the Enhanced Salp Swarm Optimization technique to select the optimal features for identifying the attacks. The proposed classifier evaluates with other different classifiers which provide a detailed study in detecting DDoS attacks using the UNSW-NB15 dataset and NSL-KDD dataset. The proposed model results 0.9936 accuracy for UNSW-NB 15 dataset and 0.9918 accuracy for NSL-KDD dataset. Empirical findings indicate that machine learning methods are highly effective at detecting and classifying attacks with a higher accuracy rate.