Multistage Optimized Fuzzy Based Intrusion Detection protocol for NIDS in MANET
Author(s):
S.Venkatasubramanian
Keywords:
Fuzzy based Framework; Mobile ad hoc networks; Hyper Parameter Optimization Techniques; NSL-KDD dataset
Abstract
Malicious activities that harm the operation of a network can be sensed using intrusion detection systems (IDS), which are critical in this regard. There are wireless networks called mobile ad hoc networks (MANETs) that don't require infrastructure to operate and can transfer data wirelessly. They are difficult to secure because of their decentralized character and the scarcity of resources. As a result, Network Intrusion Detection System (NIDS) is an excellent network security tool for detecting unknown threats in network traffic. Traditional machine learning models, such as Support Vector Machine (SVM), and others, are now the basis for most network anomaly detection systems. These methods can indeed provide some impressive results, but accuracy suffers, as they rely primarily on the manual creation of traffic features, which is no longer necessary in the age of big data This research provides a unique multi-stage optimized Fuzzy based NIDS framework to address the issues of low accuracy and feature engineering in IDS while also retaining detection performance. With the help of oversampling techniques, this research examines how small a training sample should be to get the best results. In addition, it compares and investigates the impact of detection performance and time complexity on two feature selection systems as information gain and correlation-based. It is also being explored how to improve the NIDS's performance using hyper parameter (HP) optimization methods like Random Search (RS) and Genetic Algorithm (GA). The NSL-KDD dataset is used to assess the presence of the framework. Additionally, hyper-parameter adjustment with detection accuracies improves model performance
Article Details
Unique Paper ID: 153276
Publication Volume & Issue: Volume 8, Issue 6
Page(s): 301 - 311
Article Preview & Download
Share This Article
Conference Alert
NCSST-2021
AICTE Sponsored National Conference on Smart Systems and Technologies
Last Date: 25th November 2021
SWEC- Management
LATEST INNOVATION’S AND FUTURE TRENDS IN MANAGEMENT