How radadiyaaditya/dns-filtering-with-ml works
Project Type Analysis
This is a specialized network security application that combines Software-Defined Networking (SDN) with Machine Learning to create an intelligent packet filtering system. The project aims to detect and filter unwanted traffic by analyzing domain names and their actual content.
Main Components
a) Input Layer:
• Network Traffic Interceptor (using Scapy)
• Packet Capture Component
• Domain Name Extractor
b) Processing Layer:
• Web Scraping Module (BeautifulSoup)
• ML Processing Components:
• Sentence Transformer Module
• Hugging Face Models
• Similarity Analysis Engine (simwords.py)
c) SDN Control Layer:
• SDN Controller
• Policy Enforcement Module
• Traffic Filtering Rules
d) Output Layer:
• Packet Forward/Drop Decision
• Traffic Management Actions
Specific Elements
a) Input Section:
[Network Traffic] → [Packet Capture] → [Domain Extraction]
b) Processing Section:
[Web Scraper] → [Content Analysis] → [ML Processing]
• ↓
• [Similarity Comparison Engine]
c) Decision Section:
[SDN Controller] ← [Policy Engine] ← [ML Decision Output]
• ↓
[Traffic Filtering]
Technical Details to Highlight
• Python3 as the primary implementation language
• BeautifulSoup for web content extraction
• Hugging Face and Sentence Transformers for ML processing
• Scapy for packet manipulation
• SDN integration points