All projects

gnss-toolkit

tools

GPS/GNSS testing with publication-quality plots

Python GNSS NMEA Data Analysis Matplotlib
View on GitHub

Overview

gnss-toolkit is a Python suite for characterizing GPS/GNSS receivers end to end. It captures raw NMEA output from Telit-based modules, derives the metrics that actually describe receiver performance, and renders publication-ready accuracy plots — as a CLI or a Tkinter GUI.

The problem

GNSS bring-up and validation generate streams of NMEA sentences that mean little until they’re reduced to the numbers a datasheet cares about. Doing that by hand per test is slow and inconsistent. This toolkit standardizes capture → metrics → plot so results are repeatable and comparable across antennas, receivers, and sky conditions.

What it does

Two scripts cover the workflow:

  • gnss_logger.py — resets and powers on the GNSS engine (send_reset does cold / warm / hot via AT command), enables NMEA output (enable_nmea), and logs raw sentences to file (capture_nmea, default 900 s).
  • gnss_nmea_analysis.py — parse_nmea extracts time, fix quality, lat/lon, HDOP, and satellite count; a haversine distance against a reference position yields the metrics; main writes a CSV plus the plots.

Metrics computed: TTFF (cold/warm/hot, as captured), CEP₅₀ / CEP₉₅, and RMS error. Outputs are written alongside the NMEA log as CSV + PNG figures (scatter and accuracy plots).

Local ENU scatter plot of GNSS fixes around a reference point, with CEP50 (2.24 m) and CEP95 (2.74 m) circles overlaid.
Example output: an ENU scatter of fixes with CEP₅₀ / CEP₉₅ circles around the reference position.

Tech

Python with pyserial (capture), pandas / numpy (analysis), and Matplotlib (plots); a Tkinter GUI wraps the analysis.

Get it

Source and usage are on GitHub.