How Nautilus Works

Nautilus uses a sophisticated multi-stage processing pipeline to map IP links to submarine cables, combining multiple data sources with advanced validation techniques.

Data Sources

Multiple comprehensive data sources power our submarine cable mapping system

Traceroute Data

Network path measurements showing IP links between routers across the Internet. Raw traceroute output provides the foundation for link analysis.

Sample Output:
traceroute to google.com
1 192.168.1.1
2 10.0.0.1
3 203.0.113.1
4 198.51.100.1

Cable Infrastructure

TeleGeography submarine cable database with landing points, cable routes, and ownership information for comprehensive infrastructure mapping.

Cable Info:
Cable: TAT-14
Landing Points:
- New York, USA
- London, UK
Status: Active

Geolocation Services

11 different IP geolocation providers to determine router locations with high accuracy and redundancy for validation.

Location Data:
IP: 203.0.113.1
MaxMind: London, UK
IPInfo: London, UK
GeoLite: London, UK

Ownership Mapping

4 different AS mapping services to determine infrastructure ownership and enhance cable predictions through organizational alignment.

AS Data:
IP: 203.0.113.1
Cymru: AS15169
bdrmapIT: AS15169
AS15169: Google LLC

Processing Pipeline

Advanced algorithms process and validate data through multiple stages

Link Classification

Analyzes IP links to determine if they traverse submarine cables using multiple validation techniques:

  • Geographic analysis of endpoints
  • Country-level neighbor detection
  • Land-locked country identification
  • Speed-of-light validation

Cable Mapping

Maps identified submarine links to specific cables through proximity analysis:

  • Finding nearest cable landing points
  • Calculating proximity scores
  • Recursive radius search (up to 1000km)
  • Handling multiple cable predictions

Owner-based Mapping

Enhances cable predictions using ownership data and organizational relationships:

  • AS organization name matching
  • Cable owner to IP mapping
  • Calculating ownership match scores
  • Cross-referencing with known operators

Final Mapping & Confidence Scoring

Selects cables and assigns prediction confidence based on multiple factors:

  • Geolocation accuracy consensus
  • Distance to landing points
  • AS ownership alignment with cable owners
  • Historical validation data

Validation Techniques

Multiple validation methods ensure accuracy and reliability of our predictions

Cable Failure Analysis

Links mapped to failed cables disappear during outages and return when repaired, providing real-world validation of predictions.

Targeted Traceroutes

Measurements between cable landing points validate predicted mappings and confirm infrastructure relationships.

Operator Maps

Comparison with Tata Communications and Vodafone network maps provides industry-standard validation benchmarks.

Validation Results

Our system achieves high accuracy through rigorous validation processes

Top Prediction Accuracy

77%

Exact matches

Secondary Match

19%

Alternative predictions

No Match

4%

Missed predictions

Technical Implementation

Data Processing

  • Multi-source data aggregation and normalization
  • Geographic coordinate validation and optimization
  • Real-time traceroute parsing and analysis
  • Confidence scoring algorithms (0.0-1.0 scale)

Reliability & Privacy

  • Client inputs processed server-side; no third-party sharing
  • Results include confidence scores for each inference
  • All endpoints validate input and limit payload sizes
  • Open source framework with transparent methodology