Optimized Hybrid Voting-Based Machine Learning Framework for Securing Wireless Sensor Networks Against DDoS Attacks

Dr.R.M. Mallika; R. Surekha; Maddina Keerthi; Syed Afreedh; Ellanti Harshitha; Andam Eranna Goud

Optimized Hybrid Voting-Based Machine Learning Framework for Securing Wireless Sensor Networks Against DDoS Attacks

Authors: Dr.R.M. Mallika, R. Surekha, Maddina Keerthi, Syed Afreedh, Ellanti Harshitha, Andam Eranna Goud

Unique Paper ID: 177401
Volume: 11
Issue: 12
PageNo: 484-491

Keywords: DDoS Intrusion Detection Systems(IDS) Hybrid Machine Learning Voting Classifier RF SVM Software-Defined Networks (SDN) Internet of Things (IoT).

Abstract:
Distributed Denial of Service (DDoS) attacks present significant challenges to Software-Defined Networks (SDN) and Internet of Things (IoT) environments, where dynamic traffic patterns make threat detection complex. Traditional Intrusion Detection Systems (IDS), which rely on static rules and signature-based techniques, often fail to detect evolving attack patterns, leading to high false positives and inadequate threat mitigation. This paper proposes a Hybrid Machine Learning-Based IDS that integrates Random Forest (RF) and Support Vector Machine (SVM) through a Voting Classifier to enhance detection accuracy and reduce false positives. The hybrid model leverages the strengths of RF in decision-making and SVM in boundary optimization to provide a more reliable and adaptive solution. To improve classification performance, extensive experiments were conducted using a labeled DDoS dataset (CIC-DDoS2019), applying preprocessing techniques such as feature extraction, normalization, and data augmentation to enhance data quality and balance. The hybrid model achieved 80% accuracy and demonstrated a 15% reduction in false positive rate compared to traditional IDS systems. Feature importance analysis highlighted critical indicators such as packet rate, flow duration, and source-destination interactions, which were key in distinguishing between normal and malicious traffic. The model’s adaptive learning capability and scalability make it suitable for securing modern IoT and SDN environments. The experimental results validate the system's effectiveness, and future work aims to enhance its adaptability to detect complex and evolving attack patterns, further improving real-time threat mitigation and overall network security.

Download article

email to a friend

Cite This Article

ISSN: 2349-6002
Volume: 11
Issue: 12
PageNo: 484-491

Optimized Hybrid Voting-Based Machine Learning Framework for Securing Wireless Sensor Networks Against DDoS Attacks

Available:https://ijirt.org/Article?manuscript=177401

30 Apr, 2025

Predicting Cardiac Arrest Early: A Deep Learning Approach with Statistical Models

30 Apr, 2025

ANDROID AUTOMATED MALWARE DETECTION BASED ON INFROMATION SECURITY

30 Apr, 2025

Brain Tumor Detection using transfer learning: A Breakthrough in Medical Imaging

30 Apr, 2025

MobileNet architecture -Based Pill Detection System for Accurate Drug Identification in Healthcare

30 Apr, 2025

House Price Prediction Simplified: Deep Learning in Real Estate Analytics

30 Apr, 2025

Tree Structure for Primitive Solutions of the Diophantine Equation px2 +y2- z2 = 0, p is a positive square-free integer

30 Apr, 2025

A New Approach To Generate Set Of Prime Numbers with Using Non-Overlapping and Non-Empty Sets

30 Apr, 2025

A New Approach to Generate Trees of Primitive Solutions of Diophantine Equation x2 +dy2- z2 = 0 With Using A general interval decomposition where d is a positive square free integer

30 Apr, 2025

Quantum Federated Learning Approach for Brain Tumor Classification using MRI images

Impact Factor
8.01 (Year 2024)

UGC Approved
Journal no 47859

Join Our IPN

IJIRT Partner Network

Submit your research paper and those of your network (friends, colleagues, or peers) through your IPN account, and receive 800 INR for each paper that gets published.

Join Now

Latest Publication

Recent Conferences

NCSEM 2024

National Conference on Sustainable Engineering and Management - 2024 Last Date: 15th March 2024

Submit inquiry