Review 1: A Quantitative Logarithmic Transmission-Based Network Intrusion Detection System

A Quantitative Logarithmic Transmission-based Network
Intrusion Detection System
Review 1
Internal Guide: Prof Dr. D. Ponmary Pushpa Latha Presented By:

IMRAN FARITH A – URK21DS3027
GOKUL KRISHNA J N - URK21DS3056
VIDYASAGAR S – URK21DS3022
SHYBULLAH T – URK21DS3001
Introduction
• The QLT-IDS introduces a cost-effective and efficient approach to
intrusion detection, leveraging quantitative logarithmic
transformation for network behavior analysis. This system stands
out by eliminating the extensive data collection, training
requirements, and the need for high-powered computing
resources typical of machine-learning or deep-learning based IDS
solutions.
• It effectively identifies a wide variety of malicious activities,
including both North-South and East-West attacks, crucial for
comprehensive network security. Demonstrated to perform with
high accuracy in detecting threats across real-world campus
network traffic and simulated environments, QLT-IDS proves its
capability even when analyzing significantly reduced data sets.
Relevance and Need of the Project in the
Present Context
• Efficiency and Accessibility: QLT-IDS leverages a straightforward statistical
approach, eliminating the need for extensive data training and high-end hardware.
This makes it an efficient and accessible solution for real-time, accurate threat
detection in today's cybersecurity landscape.
• Comprehensive Threat Detection: The system's unique capability to detect both

external (North-South) and internal (East-West) network threats addresses a critical
gap in traditional IDS technologies, offering comprehensive network security
coverage.
• Cost-Effective Security Solution: By reducing the reliance on intensive pretraining

and computational resources, QLT-IDS represents a cost-effective alternative to
conventional machine-learning and deep-learning-based IDS, streamlining
deployment and operational costs.
Literature Survey
Title Author Method
A QUANTITATIVE LOGARITHMIC Blue Lan The literature underscores IDS importance.

TRANSFORMATIONBASED
INTRUSION DETECTION SYSTEM
Ta-Chun QLT-IDS offers cost-effective, real-time
Rico Weth detection using simpler methods. Tested on
(2022) real and simulated traffic, it effectively
detects attacks with reduced NetFlow
records.
OVERVIEW OF INTRUSION Pierpaolo Dini IDS is vital in cybersecurity for detecting

DETECTION SYSTEMS DESIGN
EXPLOITING MACHINE LEARNING
Abdussalam Elhanashi intrusion attacks. Feature selection enhances
(2015) performance, impacted by dataset structure and
imbalance. This research explores ML
approaches using KDD 99, UNSWNB15, and
CSE-CIC-IDS 2018 datasets.
Problem Statement and Objectives
Problem Statement:
Traditional IDS: Expensive and complex, challenging to deploy and manage
effectively.
Inadequate Attack Detection: Struggles to detect various attacks, leaving
networks vulnerable.
Operational Risks: Inefficient detection leads to serious consequences like data
breaches.
Cost-effective Solutions: Demand for simpler, affordable IDS detecting threats
in real-time without extensive resources.
Problem Statement and Objectives
Objectives:
QLT-IDS Development: Create a user-friendly IDS using statistical methods for instant
intrusion detection.
Evaluating Efficacy: Test QLT-IDS on real and simulated data to ensure it detects
various attacks, including those missed by traditional systems.
Comparison with Traditional Methods: Compare QLT-IDS with traditional IDS to

determine effectiveness, efficiency, and resource needs.
Scalability and Adaptability: Assess QLT-IDS's ability to adapt to diverse networks

and scale for organizations of all sizes.
Proposed Approach
Illustrates the learning process of the model. After obtaining the dataset, the data
first needs to be pre-processed. First, data cleaning is performed, as the sample
size is sufficient, and when each sample is checked, if there are any invalid fields, the
sample is discarded. Then, due to a significant imbalance in the sample size, in this study,
5000 samples of each type were randomly selected for the subsequent experiments. Finally,
by analysing the importance of these features and selecting some of the relatively important
ones from them, they were organized into an image format in a suitable form.
Unique Features of the Proposed
System
• Quantitative Logarithmic Transformation
• Minimal Training Requirements
• Real-time Detection without High-end Hardware
• Versatility in Handling Various Attack Types
• Robustness Under High Down-sampling Rates
• Evaluation Across Real-world and Simulated

Environment
Utility value of the Proposed System
• Enhanced Security Posture

• Cost Savings
• Improved Operational Efficiency
• Risk Mitigation
• Compliance Assurance
• User-Friendly Interface
Users of the System
• Network Administrators: They use QLT-IDS for real-time threat detection and rapid
response to network security issues.
• IT Security Analysts: They leverage QLT-IDS to analyze intrusions, assess their impact, and
develop effective response strategies.
• Compliance Officers: They utilize QLT-IDS to ensure regulatory compliance and adherence
to industry standards in network security.
• System Administrators: They manage QLT-IDS performance and configuration for optimal
functionality.
• Executive Management: They rely on QLT-IDS reports and visualizations to make informed
cybersecurity decisions.
Functional Requirements
S. Description Stimul Response Dependencies and
No. us Constraints (if any)
1. Real-time Detection: Continuous monitoring and
immediate detection of network intrusions.
Multi-Protocol Support: Ability to detect intrusions across

various network protocols.
Anomaly and Rule-Based Detection: Utilization of

2. anomaly detection and customizable rule-based detection
for identifying abnormal network behaviour and known
attack signatures.
Alerting and Reporting: Generation of alerts and

3. comprehensive reports upon detecting suspicious activity,
facilitating timely response and analysis.
Integration and Configuration: Seamless integration with

4. SIEM systems and external threat intelligence feeds, along
with flexible configuration options to tailor detection
settings to organizational needs.
Non-Functional Requirements
 Performance: Handles high network traffic efficiently for real-time threat detection.
 Scalability: Expands seamlessly with network growth without performance loss.
 Reliability: Ensures high uptime and low false rates for continuous protection.
 Usability: Intuitive interface for easy operation by admins and analysts.
 Security: Adheres to strict standards for data protection.

Structure Diagram
Use Case Diagram

Review 1: A Quantitative Logarithmic Transmission-Based Network Intrusion Detection System

Uploaded by

Copyright:

Available Formats

Review 1: A Quantitative Logarithmic Transmission-Based Network Intrusion Detection System

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Review 1: A Quantitative Logarithmic Transmission-Based Network Intrusion Detection System

Uploaded by

Copyright:

Available Formats

A Quantitative Logarithmic Transmission-based Network

Intrusion Detection System

Internal Guide: Prof Dr. D. Ponmary Pushpa Latha Presented By:

• Comprehensive Threat Detection: The system's unique capability to detect both

• Cost-Effective Security Solution: By reducing the reliance on intensive pretraining

A QUANTITATIVE LOGARITHMIC Blue Lan The literature underscores IDS importance.

OVERVIEW OF INTRUSION Pierpaolo Dini IDS is vital in cybersecurity for detecting

Comparison with Traditional Methods: Compare QLT-IDS with traditional IDS to

Scalability and Adaptability: Assess QLT-IDS's ability to adapt to diverse networks

• Minimal Training Requirements

• Real-time Detection without High-end Hardware

• Versatility in Handling Various Attack Types

• Robustness Under High Down-sampling Rates

• Evaluation Across Real-world and Simulated

• Enhanced Security Posture

Multi-Protocol Support: Ability to detect intrusions across

Anomaly and Rule-Based Detection: Utilization of

Alerting and Reporting: Generation of alerts and

Integration and Configuration: Seamless integration with

 Scalability: Expands seamlessly with network growth without performance loss.

 Usability: Intuitive interface for easy operation by admins and analysts.

 Security: Adheres to strict standards for data protection.

You might also like