MULTI-LF: A Unified Continuous Learning Framework for Real-Time DDoS Detection in Multi-Environment Networks
Furqan Rustam, Islam Obaidat, Anca Delia Jurcut·April 15, 2025
Summary
MULTI-LF, a unified continuous learning framework, boosts real-time DDoS detection in multi-environment networks. It employs two machine learning models for high accuracy, low latency, and efficient resource usage. Addressing diverse malicious traffic patterns and evolving threats, MULTI-LF improves model performance through scaling, ensuring balanced weight updates and high accuracy across metrics. BNB is the preferred model in this approach. Research focuses on malware detection, network traffic analysis, and cybersecurity, covering self-paced learning, real-time monitoring, and intrusion detection frameworks. Studies explore machine learning techniques for network intrusion detection across various platforms and environments from 2010 to 2024.
Introduction
Background
Overview of DDoS attacks and their impact on multi-environment networks
Challenges in real-time detection and mitigation
Objective
To present MULTI-LF, a unified continuous learning framework designed for efficient DDoS detection
Highlighting the framework's ability to handle diverse malicious traffic patterns and evolving threats
Method
Data Collection
Techniques for gathering network traffic data in real-time
Importance of data quality and relevance for effective detection
Data Preprocessing
Methods for cleaning, normalizing, and transforming raw data
Role in enhancing model performance and accuracy
Model Selection and Training
Introduction to BNB (Bagging Nearest Neighbors) as the preferred model
Explanation of how BNB and another model work together for high accuracy, low latency, and efficient resource usage
Scaling and Weight Updates
Strategies for maintaining balanced weight updates across diverse environments
Importance in ensuring high accuracy across different metrics
Research Focus
Malware Detection and Network Traffic Analysis
Overview of research in malware detection and network traffic analysis
Importance of these areas in cybersecurity
Self-paced Learning and Real-time Monitoring
Explanation of self-paced learning in the context of MULTI-LF
Role of real-time monitoring in enhancing detection capabilities
Intrusion Detection Frameworks
Overview of intrusion detection frameworks and their evolution
How MULTI-LF contributes to these frameworks
Machine Learning Techniques for Network Intrusion Detection
Historical Overview (2010-2024)
Summary of advancements in machine learning techniques for network intrusion detection
Key milestones and influential studies
Current Applications and Future Trends
Discussion of current applications of machine learning in network intrusion detection
Insights into future trends and potential innovations
Basic info
papers
cryptography and security
machine learning
artificial intelligence
Advanced features
Insights
What innovative approaches does MULTI-LF employ to address evolving threats in network security?
In what ways does MULTI-LF ensure compatibility across different network environments?
What are the key features of the BNB model that make it preferred in the MULTI-LF framework?
How does the MULTI-LF framework integrate machine learning models to enhance DDoS detection?