Skip to content
This repository was archived by the owner on Jul 14, 2026. It is now read-only.
This repository was archived by the owner on Jul 14, 2026. It is now read-only.

Integrate TelemetrySolver as new localization algorithm #22

Description

@jonnyspicer

Summary

Integrate the TelemetrySolver (Levenberg-Marquardt TDOA/FDOA solver) into 3lips as a new localization algorithm option alongside the existing EllipseParametric, EllipsoidParametric, and SphericalIntersection methods.

Background

The TelemetrySolver provides:

  • Levenberg-Marquardt optimization for TDOA/FDOA localization
  • Velocity estimation in addition to position
  • Robust convergence on bistatic radar measurements
  • Input: 3 simultaneous detections with bistatic range and Doppler
  • Output: Single position (LLA) with velocity (ENU)

Implementation Plan

Phase 1: Module Integration (2-3 hours)

  • Create TelemetrySolverLocalisation.py wrapper class
  • Import TelemetrySolver modules directly (no subprocess calls)
  • Add to localization algorithm selection in event.py
  • Update Docker configuration to include TelemetrySolver

Phase 2: Testing & Validation (1 hour)

  • Test module imports work correctly
  • Verify data format conversion from 3lips to TelemetrySolver
  • Test with synthetic-adsb data
  • Manual verification on map interface

Phase 3: Real Radar Testing

  • Test with real radar units
  • Compare results with existing localization methods
  • Verify positions match ADS-B truth data

Technical Details

Input Format: Same as existing methods (delay in km, Doppler in Hz)
Output Format: Standard 3lips localization format with points array
Dependencies: numpy, scipy (already in 3lips)
API Usage: "localisation": "telemetry-solver"

Files to Modify

  • 3lips/event/algorithm/localisation/TelemetrySolverLocalisation.py (new)
  • 3lips/event/event.py (add import and registration)
  • 3lips/event/Dockerfile (copy TelemetrySolver directory)

Success Criteria

  • TelemetrySolver standalone functionality verified
  • Module imports work without errors
  • Positions appear reasonable on map interface
  • Can manually verify positions match expected aircraft locations
  • Integration doesn't break existing functionality

Priority

High - needed for real radar testing

Time Estimate

~3-4 hours total for MVP integration ready for real radar testing

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

mvpMVP-critical issue

Type

No type

Projects

No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions