By Peter Steigenberger, Steffen Thoelert, Oliver Montenbruck and Richard B. Langley

The first GPS III satellite ,“Vespucci,” was launched in December 2018, started signal transmission in January 2020, and was set healthy later that month. The second GPS III satellite, nicknamed “Magellan,” was launched on Aug. 22, 2019, on a Delta IV rocket from Cape Canaveral, Florida.

Magellan, also identified by its space vehicle number (SVN) 75 (here referred to as GPS-75), started signal transmission with standard pseudorandom noise code (PRN) number 18 (here referred to as G18) on March 13. The L1 C/A, L1 P(Y), and L2 P(Y) signals were activated at 17:16:30 GPS Time (GPST), while the L1C, L2C and L5 signals followed less than two hours after Vespucci’s launch at 18:59:30 GPST. Transmission of navigation messages started at 19:00:00 GPST with GPS-75 (G18) marked as unhealthy.

PRN G18 was previously used by the 27-year-old Block IIA satellite GPS-34 that had been already removed from the active GPS constellation on Oct. 7, 2019, but continued signal transmission until March 9, 2020. GPS-75 is already being tracked by a large number of tracking stations of the International GNSS Service (IGS). Based on the data collected by these stations, the Center for Orbit Determination in Europe (CODE), headquartered in Bern, Switzerland, has been providing precise orbit and clock products for this satellite since March 14.

A comparison we performed with the CODE precise orbit products revealed initial broadcast ephemeris errors of up to 100 meters (3D) and an orbit-related signal-in-space range error (SISRE) of about 13 meters. Within four days, a SISRE (orbit component) of 24 centimeters was achieved, which closely matches the performance of the rest of the GPS constellation.

Figure 1 shows the spectral flux density of GPS-75 in the L1, L2 and L5 frequency bands obtained with the 30-meter high-gain antenna of the German Aerospace Center (DLR) located in Weilheim, Germany. The civil L1 C/A, L1C and L2C signals can be identified as sharp peaks in the center of the respective frequency bands.

FIGURE 1. Spectral flux density of GPS-75 measured with DLR’s 30-meter high-gain antenna. (Figure: Steigenberger, et al)

The prominent side lobes in the L1 and L2 bands are associated with the military M-code. The wide main lobe of the L5 signal with two smaller and sharper side lobes is caused by the superposition of two in-phase and quadrature signals with a 10-MHz binary phase-shift keying (BPSK) modulation. We found that all signals are in good shape and have a quality similar to that of the first GPS III satellite.

On March 16, 2020, we detected a significant change in the carrier-to-noise-density ratio of the L1 and L2 P(Y)-code signals. Figure 2 illustrates these changes for the IGS station located in Patumwan, Thailand (CUSV00THA). The L1 and L2 P-code signals are usually encrypted with the W-code to prevent spoofing (the generation of fake signals by adverse parties). The resulting encrypted signals are denoted by P(Y). Geodetic GNSS receivers are capable of tracking the P(Y) signals with a semi-codeless approach.

FIGURE 2. Carrier-to-noise-density ratio (C/N₀) of the second GPS III satellite, GPS-75, tracked by the IGS station CUSV00THA in Patumwan, Thailand, on March 16, 2020. Between 20:22 and 21:18 GPST, unencrypted P-code signals were tracked. (Figure: Steigenberger, et al)

As a result, C/N₀ of L1 P(Y) and L2 P(Y) are virtually identical and significantly smaller than the C/N₀ of the unencrypted signals due to losses of the semi-codeless tracking technique. This can be seen in the blue-colored plot of Figure 2, where the C/N₀ values of L1 P(Y) and L2 P(Y) are identical and smaller by 4.5–16 dB compared to L1 C/A depending on the elevation angle of the satellite.

However, between 20:22 and 21:18 GPST, an increase of the P-code C/N₀ values was observed. The values changed by 4.5 and 12.5 dB for L1 and L2, respectively. This change is an indicator that unencrypted P-code signals were transmitted, rather than encrypted ones. This assumption can be verified by the “Anti-Spoof Flag” given as the 19th bit of the handover word (HOW) of the GPS LNAV navigation message.

Indeed, decoding of the raw navigation data from the IGS station CHOF00JPN in Chofu, Japan, showed that the Anti-Spoof Flag indicated a deactivation of anti-spoofing between 20:22:00 and 21:17:48 GPST and verified our assumption that unencrypted P-code signals were transmitted during that time period.

It has to be noted that only Javad receivers within the global multi-GNSS network of the IGS show this increase in C/N₀. Other receiver types report continuous C/N₀ values for the P-code signals, indicating that a semi-codeless tracking technique was continuously applied irrespective of the Anti-Spoof Flag.

Figure 3 shows the two GPS III satellites’ Allan deviation, which measures the clock stability achieved in orbit; that is, the average frequency error…