RIPE Atlas Internet Path Analytics: Global Traceroute Enrichment

This repository features a comprehensive Python-based data pipeline designed to analyze and visualize internet routing paths. Using RIPE Atlas traceroute data targeting Anchor 6360 in Salzburg, Austria (AS39878), this project explores the complexities of global internet topology, autonomous system (AS) transitions, and geolocation bottlenecks.

Developed as a collaborative seminar project at the University of Vienna, this toolset transforms raw, multi-gigabyte JSONL traceroute dumps into enriched, actionable network insights.

Key Features

• Streaming Data Pipeline: Efficiently processes large-scale .jsonl and .bz2 datasets without high memory overhead.
• Enrichment Engine: Automatically enriches traceroute hops with:
  • Reverse DNS (PTR): Resolving hostnames for infrastructure identification.
  • ASN Mapping: Identifying Autonomous Systems via RIPEstat and CAIDA datasets.
  • Geolocation: Mapping unique IP addresses to cities and countries using MaxMind and IPinfo APIs.
• Smart Caching: Implements a local caching mechanism to minimize redundant external API lookups, ensuring speed and cost-effectiveness.
• Advanced Visualizations: Includes automated heatmaps of country-to-country transitions and graph-based representations of transit provider dependencies.

Technical Results & Insights

The analysis of 88,302,907 traceroute measurements revealed a high dependency on major transit providers (AS1299 - Telia, AS6939 - Hurricane Electric) for traffic destined for Austria.

Core Statistics

• Unique IPs Observed: 1,004
• Resolved PTR Records: 600
• Top Transit ASN: AS1299 (92 unique paths)
• Global Reach: Traceroutes originated from over 40+ countries, with significant transit clusters in DE, US, and NL.

Repository Structure

• parser.ipynb: Filters and simplifies raw RIPE Atlas measurements into a manageable format.
• enrichment.ipynb: The core logic for streaming IP enrichment and geolocation caching.
• geomap.ipynb: Generates interactive HTML world maps of unique IP distributions.
• visualisation1.ipynb: Creates graph networks of frequent transit Autonomous Systems.
• fns.pdf: The complete technical seminar report detailing the research methodology and results.

Build and Execution

Prerequisites

• Python 3.10+
• Pandas, Plotly, GeoPandas, NetworkX, Matplotlib

Installation

pip install pandas plotly geopandas networkx matplotlib requests

Usage Run parser.ipynb to process the raw .bz2 or .jsonl RIPE Atlas data.

Execute enrichment.ipynb to perform the IP-to-ASN and Geolocation lookups.

Use the visualisation notebooks to generate performance heatmaps and topological graphs.

Collaborators: Mihajlo Katić, Jovana Gojković, Julian Waluschyk University of Vienna - Computer Science