Skip to content

Quick Start: MADOCA-PPP

This guide walks you through running MADOCA-PPP (Precise Point Positioning) with MRTKLIB, covering both basic PPP and PPP with ambiguity resolution (PPP-AR).

MADOCA (Multi-GNSS Advanced Demonstration tool for Orbit and Clock Analysis) provides real-time satellite orbit, clock, and bias corrections via QZSS L6E. Combined with broadcast ephemeris, it enables decimeter-to-centimeter-level global PPP positioning.

Prerequisites

  • MRTKLIB built successfully (cmake --build build)
  • Test data extracted:
Bash
tar xzf tests/data/madocalib/madocalib_testdata.tar.gz -C tests/data/madocalib/

Required Files

File TOML Key Purpose
RINEX observation (.obs) positional arg GNSS observations from a receiver
RINEX navigation (.rnx / .nav) positional arg Broadcast ephemeris (mixed GNSS)
L6E correction (.l6) positional arg MADOCA SSR corrections (orbit, clock, bias)
Antenna model (.atx) files.receiver_atx Receiver/satellite antenna phase center model

ANTEX File: IGS20 for MADOCA

Unlike CLAS (which requires IGS14), MADOCA-PPP uses the IGS20/ITRF2020 reference frame. Use igs20.atx for antenna corrections.

Optional Files (for PPP-AR)

For ambiguity-resolved PPP, no additional files are required beyond the L6E corrections — MADOCA provides the necessary phase bias information via SSR.

Bundled Configuration Files

MRTKLIB provides three MADOCA-PPP configurations with increasing capability:

File Mode AR Iono Frequencies
conf/madocalib/rnx2rtkp.toml PPP Off dual-freq L1+L2
conf/madocalib/rnx2rtkp_pppar.toml PPP-AR Continuous est-stec L1+L2+L3+L4
conf/madocalib/rnx2rtkp_pppar_iono.toml PPP-AR + Iono Continuous est-stec L1+L2+L3+L4

Post-Processing

Basic PPP (Float Solution)

The simplest mode — no ambiguity resolution, dual-frequency ionospheric correction:

Bash
mrtk post -k conf/madocalib/rnx2rtkp.toml \
  -ts 2025/04/01 0:00:00 -te 2025/04/01 0:59:30 -ti 30 \
  -o output_madoca_ppp.pos \
  tests/data/madocalib/MIZU00JPN_R_20250910000_01D_30S_MO.obs \
  tests/data/madocalib/BRDM00DLR_S_20250910000_01D_MN.rnx \
  tests/data/madocalib/2025091A.204.l6 \
  tests/data/madocalib/2025091A.206.l6

Multiple L6E Files

MADOCA uses two L6E streams (mdc-003 and mdc-004) on different PRNs. Pass both .l6 files as positional arguments. The decoder selects the appropriate stream automatically.

Expected accuracy: Decimeter-level (Q=5 Float).

PPP-AR (Ambiguity Resolution)

For centimeter-level accuracy with integer ambiguity resolution:

Bash
mrtk post -k conf/madocalib/rnx2rtkp_pppar.toml \
  -ts 2025/04/01 0:00:00 -te 2025/04/01 0:59:30 -ti 30 \
  -o output_madoca_pppar.pos \
  tests/data/madocalib/MIZU00JPN_R_20250910000_01D_30S_MO.obs \
  tests/data/madocalib/BRDM00DLR_S_20250910000_01D_MN.rnx \
  tests/data/madocalib/2025091A.204.l6 \
  tests/data/madocalib/2025091A.206.l6

Key differences from basic PPP:

  • frequency = "l1+2+3+4" — Multi-frequency processing
  • ambiguity_resolution.mode = "continuous" — AR enabled
  • ionosphere = "est-stec" — Slant TEC estimation (better than dual-freq for AR)
  • systems includes all 5 constellations (GPS, GLONASS, Galileo, QZSS, BeiDou)

Expected accuracy: Centimeter-level after convergence (Q=1 Fix).

PPP-AR with Ionospheric Correction

The highest-accuracy mode, using MADOCA ionospheric corrections:

Bash
mrtk post -k conf/madocalib/rnx2rtkp_pppar_iono.toml \
  -ts 2025/04/01 0:00:00 -te 2025/04/01 0:59:30 -ti 30 \
  -o output_madoca_pppar_iono.pos \
  tests/data/madocalib/MIZU00JPN_R_20250910000_01D_30S_MO.obs \
  tests/data/madocalib/BRDM00DLR_S_20250910000_01D_MN.rnx \
  tests/data/madocalib/2025091A.204.l6 \
  tests/data/madocalib/2025091A.206.l6

This adds receiver.iono_correction = true, enabling MADOCA-specific ionospheric corrections for faster convergence.

Key Configuration Points

Ephemeris Source

TOML
[positioning]
satellite_ephemeris = "brdc+ssrapc"

MADOCA provides SSR corrections that are applied to broadcast ephemeris. brdc+ssrapc uses antenna phase center-corrected SSR orbits.

Signal Selection

TOML
[signals]
gps = "L1/L2/L5"
qzs = "L1/L5/L2"
galileo = "E1/E5a/E5b/E6"
bds2 = "B1I/B3I/B2I"
bds3 = "B1I/B3I/B2a"

The [signals] section controls which frequency pairs are used for each constellation. These settings affect which observations enter the PPP filter and must match the available MADOCA bias corrections.

AR Thresholds (PPP-AR)

TOML
[ambiguity_resolution.thresholds]
ratio = 1.5     # LAMBDA validation ratio
ratio1 = 1.5    # Max 3D position std-dev to start narrow-lane AR (m)

ratio1 gates AR activation — narrow-lane integer search only begins when the float solution's 3D position standard deviation drops below this value. This prevents premature AR attempts during initial convergence.

Measurement Error Model

TOML
[kalman_filter.measurement_error]
code_phase_ratio_L1 = 300.0
code_phase_ratio_L2 = 300.0
ura_ratio = 0.1
  • Code/phase ratio of 300 (vs 100 for RTK) reflects PPP's heavier reliance on carrier phase measurements.
  • URA ratio scales satellite weighting by broadcast User Range Accuracy.

Comparison: MADOCA-PPP vs CLAS PPP-RTK

MADOCA-PPP CLAS PPP-RTK
Coverage Global Japan only
Signal QZSS L6E QZSS L6D
Convergence 10-30 minutes 1-2 minutes
Accuracy (converged) cm-level (PPP-AR) cm-level
Reference frame ITRF2020 (IGS20) ITRF2014 (IGS14)
ANTEX file igs20.atx igs14.atx
Grid files needed No Yes (Japan grid)
ISB table needed No Yes

Troubleshooting

No Convergence (Position Jumps)

  1. Wrong ANTEX — Ensure igs20.atx is used (not igs14.atx).
  2. Missing L6E data — Verify both L6E files are passed as arguments.
  3. Elevation mask too high — 10 degrees is standard for PPP; higher values reduce satellite count and slow convergence.

PPP-AR Not Fixing

  1. ratio1 too tight — If the float solution is noisy, AR won't activate. Try increasing ratio1 to 2.0-3.0 m.
  2. Short observation span — PPP-AR typically needs 15-30 minutes of data to converge and resolve ambiguities.
  3. Constellation limitation — Ensure systems includes all available constellations for maximum satellite geometry.

Next Steps