GPS Toolbox

Toolbox for static GPS positioning using RINEX/SP3 data and single-station or double-difference workflows

GPS Toolbox is a MATLAB project for static GPS positioning that integrates RINEX/SP3/atmospheric processing with both single-station and two-station relative-positioning workflows. It supports GPS RINEX 2.x and 3.x observation and navigation files, including automatically decompressed Hatanaka CRINEX (.crx) observations. Mixed multi-GNSS broadcast-navigation files may be read, although only GPS ephemerides and GPSA/GPSB Klobuchar ionospheric parameters are retained. Precise processing preferentially uses a local IGS SP3 orbit file when available; otherwise, broadcast ephemerides are used. Local ANTEX corrections support receiver PCO/PCV and satellite APC with nadir-dependent PCV, while BLQ ocean-loading corrections are applied only when coefficients are available for both stations, preventing asymmetric loading corrections.

For standalone positioning, the toolbox estimates a static receiver’s WGS-84 position and a GPS-time receiver-clock bias at each epoch from GPS code observations and broadcast ephemerides. It performs iterative weighted least-squares estimation epoch by epoch, followed by a robust joint static adjustment that estimates one common receiver position and independent receiver-clock biases for the valid epochs. With dual-frequency data, it can form ionosphere-free code and carrier-phase combinations and run parallel broadcast-orbit and precise-orbit/SP3 code solutions. In the short-session reported solution, the horizontal coordinate is retained from the RINEX APPROX POSITION XYZ header, and the vertical correction is obtained by inverse-variance fusion of the broadcast and precise-clock SP3 code solutions.

The toolbox also provides static double-difference (DD) relative positioning for two GPS stations. Core RTKLIB 2.4.3 algorithms—undifferenced residual computation (zdres), common-satellite selection (selsat), DD residual formation (ddres), and DD covariance construction (ddcov)—have been ported into project-root MATLAB .mfiles and integrated with the existing RINEX, SP3, and atmospheric stack. The default configuration processes rover AMC4 against base NIST over an approximately 147 km baseline, using the bundled one-hour dataset sampled every 30 seconds for 120 epochs. NIST is normally held fixed at its RINEX APPROX POSITION XYZ during DD weighted least squares, although optional single-station base refinement is available but disabled by default. AMC4 is estimated through batch ionosphere-free code-and-phase DD float weighted least squares using the RTKLIB IONOOPT_IFLCambiguity model. After the DD solution, a free-network datum translation distributes the remaining translation between the two RINEX approximate station coordinates while preserving the estimated DD baseline; the toolbox then reports absolute coordinates, baseline length, and baseline ENU components.