BLCD Real-Time Intrusion Detection System
BLCD is a Python-based Real-Time Intrusion Detection System designed to monitor live network traffic, extract security-relevant features, classify suspicious activity with machine learning, and surface incidents through alerts, persistent logs and a SOC-style dashboard.
Problem Statement
Security teams need visibility into network activity as it happens. Static analysis or post-incident logs can delay detection, while raw packet data is too noisy to interpret directly. BLCD connects packet capture, feature extraction, ML classification and analyst-facing alerting into one workflow.
Objective
Build a practical real-time IDS that captures traffic, converts packets into machine-learning features, classifies attack categories, records incidents and gives an analyst an immediate operational view.
System Architecture
- 01→Live Network Traffic
- 02→Packet Capture with Scapy
- 03→Feature Extraction with PyShark / Tshark
- 04→Dataset Creation and Preprocessing
- 05→RandomForest Model Training
- 06→Real-Time Attack Classification
- 07→Alerts + Persistent Logs
- 08→SOC-Style Streamlit Dashboard
Dataset
Project dataset pipeline built from captured/combined traffic CSV data.
Attack classes documented in the project include Normal traffic, Port scanning, SYN flood and ICMP flood.