Paper Key : IRJ************766
Author: Maneesh K M,Akash Shamrao Doltade ,Krishnendhu P G,Lakshmika Unnikrishnan ,Linny Sunny
Date Published: 10 Nov 2024
Abstract
In wireless sensor networks (WSNs), Intrusion Detection Systems (IDS) play a crucial role in identifying and mitigating unknown or unidentified attacks that aim to disrupt or compromise the networks functionality. These attacks may target the networks ability to collect, process, and transmit data, thereby preventing it from performing its intended tasks. The proposed system leverages Convolutional Neural Networks (CNNs), a powerful deep learning technique, combined with ANOVA (Analysis of Variance) feature selection to detect such attacks. CNNs are utilized to automatically learn complex patterns in the network traffic data, making it highly effective in identifying both known and unknown attacks. Meanwhile, ANOVA feature selection is employed to identify the most significant features from the sensor data that contribute to accurate attack detection, improving the systems efficiency by reducing computational overhead and focusing on the most relevant data points. This integrated approach aims to enhance the intrusion detection accuracy and network security by effectively distinguishing between normal network behaviors and malicious activities.