Cloud-Based Network Traffic Monitoring and Anomaly Detection for Healthcare Infrastructure

🩺 Project Overview

This project provides a scalable, Dockerized monitoring system designed for cloud-based healthcare infrastructures. It simulates a multi-tier hospital network and collects real-time medical metrics (SpO2, heart rate, ECG, etc.) from patient devices. The system is deployed on AWS and leverages Prometheus, Grafana, and Alertmanager to visualize and alert on anomalies.

🔍 Focused on ensuring reliability, real-time observability, and early anomaly detection within hospital networks using cloud-native tooling.

---

⚙️ Features

• 🚑 Healthcare-specific metric simulation: Heart rate, SpO2, BP, etc.
• ☁️ Cloud-native deployment on AWS EC2 using Docker.
• 📈 Prometheus-based metric scraping with 5s interval.
• 📊 Grafana dashboards for visualization (per ward, department, patient).
• 🚨 Alertmanager with threshold-based anomaly detection.
• 🧠 ML-based anomaly detection pipeline for vitals.

---

🧱 Project Structure

CN_Project/
│
├── services/                   # All application services
│   ├── main_host/              # Central metrics collection service
│   ├── ml_service/             # Anomaly detection service
│   └── patient_simulator/      # Patient data simulation
│
├── config/                     # All configuration files
│   ├── prometheus/             # Prometheus and alerting config
│   ├── alertmanager/           # Alert routing configuration
│   ├── grafana/                # Grafana dashboards and datasources
│   └── environment/            # Environment-specific configs
│
├── data/                       # Data storage
│   └── patient_samples/        # Patient data samples
│
├── scripts/                    # Utility scripts
│   ├── setup.sh                # Setup script
│   └── deploy_aws.sh           # AWS deployment script
│
├── docs/                       # Documentation
│   ├── Images/                 # Documentation images
│   ├── AWS_Deployment_Report.md
│   └── API.md                  # API documentation
│
├── docker-compose.yml          # Main docker-compose file
└── README.md

---

🚀 Deployment Instructions

Prerequisites

• Docker & Docker Compose
• AWS EC2 (Ubuntu)
• Python 3.9+

Step-by-Step Guide

1. Clone the repository

   git clone https://github.com/KMohnishM/CN_Project.git
   cd CN_Project

2. Run the setup script

   chmod +x scripts/setup.sh
   ./scripts/setup.sh

3. Launch the system

   docker-compose up --build

4. Access interfaces

   • Prometheus: http://localhost:9090
   • Grafana: http://localhost:3001
   • AlertManager: http://localhost:9093

---

📊 Dashboards

• Real-time ECG, HR, and SpO2 dashboards
• Department-wise heatmaps
• Alert timelines and incident patterns

---

⚠️ Alerting Rules (Prometheus)

groups:
  - name: hospital-vital-alerts
    rules:
      - alert: LowSpO2
        expr: spo2_percent < 90
        for: 1m
        labels:
          severity: critical
        annotations:
          summary: "Low SpO2 for patient {{ $labels.patient_id }}"
          description: "SpO2 below 90% for more than 1 minute."

---

🧪 Simulated Metrics

• Heart Rate (BPM)
• Blood Pressure (Systolic/Diastolic)
• SpO2 (%)
• Respiratory Rate
• ECG waveform (synthetic)
• FiO2 (%), EtCO2, WBC Count, Glucose, Lactate

---

📦 Future Improvements

• Enhanced ML-based anomaly detection
• Consolidated web dashboard interface
• Kubernetes-based orchestration
• Federated monitoring for multi-hospital use
• Real-time collaboration and logging pipeline

---

🧠 Author

Kodukulla Mohnish Mythreya
2nd Year CSE, VIT Vellore
LinkedIn | GitHub

---

📜 License

This project is open-sourced under the MIT License. See LICENSE for more details.