Chapter 18: Global Navigation Satellite Systems¶
Chapter Overview Global Navigation Satellite Systems (GNSS) are space-based radio positioning services that have become foundational infrastructure for transportation, surveying, telecommunications, scientific research, and national security. This chapter presents the physics of satellite ranging, signal structure and error modelling, the four global constellations (GPS, GLONASS, Galileo, BeiDou), regional systems, augmentation architectures, high-precision techniques (RTK, PPP), timing services, and the growing landscape of jamming and spoofing threats.
18.1 Fundamentals of Satellite Ranging¶
18.1.1 Pseudorange Measurement¶
A GNSS receiver estimates the signal travel time \(\Delta t\) from satellite to receiver and computes a pseudorange:
where \(c = 2.998 \times 10^8\) m s\(^{-1}\) is the speed of light, \(\hat{t}_u\) and \(\hat{t}_s\) are receiver and satellite clock readings, \(\delta t_u\) and \(\delta t_s\) are their respective clock errors relative to GNSS system time, \(I\) and \(T\) are ionospheric and tropospheric delays, and \(\varepsilon\) encompasses multipath and receiver noise. The geometric range is \(\|\mathbf{r}_u - \mathbf{r}_s\|\), the Euclidean distance between the unknown receiver position \(\mathbf{r}_u\) and the known satellite position \(\mathbf{r}_s\).
18.1.2 Position Solution: Linearisation and Least Squares¶
Four unknowns exist: \(x_u, y_u, z_u\) (receiver position in ECEF) and \(\delta t_u\) (receiver clock bias). With \(n \ge 4\) pseudorange observations, the system is linearised about an approximate solution \(\mathbf{r}_0\):
where \(\mathbf{h}_k = -(\mathbf{r}_s^k - \mathbf{r}_0)/\|\mathbf{r}_s^k - \mathbf{r}_0\|\) is the unit vector from the approximate position to satellite \(k\) and \(\delta\mathbf{x}\) is the correction to be estimated. In matrix form:
18.1.3 Dilution of Precision¶
The covariance of the position/time solution is \(\mathbf{Q} = \sigma_{\rho}^2 (\mathbf{H}^T \mathbf{H})^{-1}\). Dilution of Precision (DOP) factors quantify the geometric amplification of ranging error:
Optimal satellite geometry (satellites spread uniformly across the sky) yields PDOP \(\approx 1.5\). PDOP \(>6\) is considered poor. The User Range Error (URE) for GPS SPS is specified at \(<2.9\) m (95%) (Department of Defense 2020).
18.2 GPS (NAVSTAR Global Positioning System)¶
18.2.1 Constellation Configuration¶
The GPS constellation (Block IIR/IIR-M/IIF/III) consists of a minimum of 24 operational satellites (plus 3–6 active spares) in 6 orbital planes (\(\Delta\Omega = 60\degree{}\) spacing) with 4 satellites per plane:
-
Orbital altitude: 20,200 km (semi-major axis \(a = 26,560\) km)
-
Inclination: \(55\degree{}\)
-
Period: 11 h 58 min (approximately 2 sidereal revolutions/day)
-
Ground repeat: sidereal day (geometry repeats daily, 4-min earlier)
18.2.2 GPS Signal Structure¶
GPS signals are generated by a master oscillator at \(f_0 = 10.23\) MHz. All signal frequencies are coherent multiples:
| Signal | Frequency (MHz) | Code | Chip Rate (Mcps) | Modulation |
|---|---|---|---|---|
| L1 C/A | 1575.42 | Gold code (PRN), 1023 chips | 1.023 | BPSK-R(1) |
| L1 P(Y) | 1575.42 | P-code (encrypted Y) | 10.23 | BPSK-R(10) |
| L1C | 1575.42 | BOC(1,1) + TMBOC | 1.023/10.23 | TMBOC |
| L2C (CM+CL) | 1227.60 | 767 + 75372 chips | 0.5115 | BPSK-R(0.5) |
| L5 (I5+Q5) | 1176.45 | Neumann–Hofman | 10.23 | BPSK-R(10) |
| M-code | L1, L2 | Classified | 5.115 | BOC(10,5) |
The L1 C/A code repeats every 1 ms (1023 chips at 1.023 Mcps). Each satellite uses a unique PRN code (Gold code), enabling code-division multiple access (CDMA). The navigation message (50 bps on L1 C/A) contains broadcast ephemeris parameters, satellite clock corrections (polynomial coefficients \(a_{f0}, a_{f1}, a_{f2}\)), ionospheric model (Klobuchar coefficients \(\alpha_i, \beta_i\)), and almanac.
18.2.3 Navigation Message and Broadcast Ephemeris¶
Satellite positions are computed from Keplerian elements plus 9 perturbation parameters:
Satellite clock error is modelled as:
The relativistic correction \(\Delta t_{rel} = -\frac{2\sqrt{\mu a}}{c^2} e \sin E\) (\(\approx \pm 70\) ns) accounts for the eccentricity-dependent radial motion.
18.3 GNSS Error Budget¶
18.3.1 Ionospheric Delay¶
The ionosphere (60–1000 km altitude) introduces a dispersive delay proportional to the total electron content (TEC) along the signal path:
where \(\nu\) is the signal frequency in Hz and TEC is in TEC units (1 TECU \(= 10^{16}\) el m\(^{-2}\)). Typical TEC values: 5–30 TECU, giving \(I_{L1} \approx 1\)–\(15\) m. For dual-frequency receivers, the ionosphere-free (IF) combination eliminates the first-order term:
Single-frequency users apply the Klobuchar broadcast model which corrects \(\approx 50\%\) of the ionospheric error.
18.3.2 Tropospheric Delay¶
The neutral atmosphere delay is non-dispersive (same for all GNSS frequencies). Total tropospheric delay at zenith is approximately 2.3 m (dry component 2.1 m, wet component 0.1–0.4 m). The mapping function \(m(\varepsilon)\) scales zenith delay to elevation angle \(\varepsilon\):
The Saastamoinen zenith dry delay model (Saastamoinen, J. 1973):
where \(P_0\) is surface pressure (hPa), \(\phi\) is latitude, and \(H\) is height (km).
18.3.3 Relativistic Effects¶
Two relativistic effects apply to GPS clocks (Ashby, N. 2003):
-
Special relativity (time dilation): Satellite clock runs \(-7.2\,\mu\)s/day slow due to orbital velocity (\(v_s \approx 3.87\) km/s).
-
General relativity (gravitational blue shift): Satellite clock runs \(+45.9\,\mu\)s/day fast due to lower gravitational potential at 20,200 km.
Net effect: \(+38.4\,\mu\)s/day. GPS clocks are factory-offset to \(f_0^* = f_0 \times (1 - 4.464 \times 10^{-10})\) to cancel this.
18.3.4 Error Budget Summary¶
| Error Source | Typical Value (m) | Mitigation |
|---|---|---|
| Satellite clock (broadcast) | 0.5–1.0 | Broadcast corrections |
| Satellite orbit (broadcast) | 1.0–2.0 | Broadcast ephemeris |
| Ionosphere (Klobuchar) | 2.0–5.0 | Dual freq or model |
| Troposphere (model) | 0.5–1.0 | Saastamoinen model |
| Multipath | 0.5–5.0 | Choke ring, CRPA |
| Receiver noise | 0.1–1.0 | Code smoothing |
| Typical UERE (RMS) | 3–8 | — |
18.4 Other Global GNSS Constellations¶
18.4.1 Galileo¶
Galileo is the European Union’s civil GNSS, operated by the EU Agency for the Space Programme (EUSPA). As of 2024, 28 satellites are operational in 3 MEO orbital planes:
-
Altitude: 23,222 km; Inclination: \(56\degree{}\); Period: \(\sim 14\) h 4 min
-
Frequencies: E1 (1575.42 MHz, interoperable with GPS L1), E5a (1176.45 MHz \(=\) L5), E5b (1207.14 MHz), E6 (1278.75 MHz)
Galileo services: Open Service (OS, free), Commercial Service (CS, authenticated high-accuracy), Public Regulated Service (PRS, encrypted, government), Safety-of-Life (SoL), SAR/Galileo (MEOSAR Galileo contribution). Navigation Message Authentication (OSNMA) is a unique security feature providing open-service signal authentication to resist spoofing (Fernandez-Hernandez, I., et al. 2016).
18.4.2 GLONASS¶
Russia’s GLONASS constellation uses frequency-division multiple access (FDMA) for its legacy signals, in which each satellite transmits on its own frequency channel \(k\) (the channel numbers are re-used by antipodal satellites):
Code-division (CDMA) signals were added with the modernized GLONASS-K satellites: the first GLONASS-K1 satellite, launched in 2011, introduced the L3OC signal at 1202.025 MHz, and the later GLONASS-K2 generation adds CDMA signals on L1 and L2 as well, interoperable with other constellations. The constellation consists of 24 satellites in 3 orbital planes at 64.8 inclination and 19,100 km altitude.
18.4.3 BeiDou Navigation Satellite System (BDS-3)¶
BeiDou-3 (fully operational 2020) is China’s global GNSS with a unique mixed-orbit architecture:
-
24 MEO (Walker 24/3/1 at 21,528 km, 55) + 3 IGSO (35,786 km inclined geosynchronous) + 3 GEO (35,786 km, 80E, 110.5E, 140E)
-
Frequencies: B1C (1575.42 MHz, compatible with L1/E1), B2a (1176.45 MHz, compatible with L5), B2b (1207.14 MHz), B3I (1268.52 MHz)
BDS-3 innovations: global Short Message Communication (SMC) via L-band, Precise Point Positioning service (PPP-B2b, 0.1 m level in China), satellite-based search and rescue (MEOSAR), and global coverage including polar regions via MEO satellites (Yang, Y., et al. 2019).
18.4.4 Regional Systems¶
QZSS (Quasi-Zenith Satellite System, Japan): 4 operational satellites (1 GEO + 3 QZSS orbit, 44N/S inclination) providing GPS augmentation over Japan and Asia-Oceania with \(>3\) satellites at high elevation simultaneously. L1C/A, L1C, L2C, L5, LEX (L6) signals.
NavIC/IRNSS (India): 7-satellite constellation (3 GEO + 4 IGSO) covering India and 1500 km surroundings. L5 (1176.45 MHz) and S-band (2492.028 MHz) dual-frequency.
18.5 Differential GNSS and SBAS¶
18.5.1 Differential GNSS¶
A reference station with precisely known coordinates computes pseudorange corrections:
These corrections are broadcast to rover receivers, which subtract them from their own measurements. Errors correlated in space and time (satellite clock, orbit, ionosphere for short baselines \(<\) 50 km) are largely cancelled, achieving sub-metre accuracy in code-differential mode and centimetre accuracy with carrier-phase.
18.5.2 Satellite-Based Augmentation Systems (SBAS)¶
SBAS provides integrity and differential corrections via GEO satellite broadcasts:
| System | Region | GEO Satellites | LPV-200 Capability |
|---|---|---|---|
| WAAS | North America | 3 (Inmarsat, ANIK) | Yes |
| EGNOS | Europe | 3 (Inmarsat, SES) | Yes |
| MSAS | Japan | 2 (MTSAT) | Limited |
| GAGAN | India | 2 (GSAT) | Yes |
| SDCM | Russia | 2 (Luch-5) | No |
| BDSBAS | China | 3 (BDS GEO) | No |
SBAS broadcasts fast corrections (FCs, for satellite clock errors, updated every 6 s), slow corrections (LTCs, for orbit errors, updated every 120 s), and integrity parameters (User Differential Range Error, UDRE; Grid Ionospheric Vertical Error, GIVE). LPV-200 (Localizer Performance with Vertical guidance, 200 ft DA) requires \(\sigma_{\text{VPL}} < 35\) m and horizontal protection level \(< 40\) m (RTCA 2020).
18.6 Real-Time Kinematic (RTK)¶
18.6.1 Carrier-Phase Measurement¶
The carrier-phase measurement on frequency \(j\) is:
where \(N_j\) is the integer ambiguity (cycles). Differencing between satellites (single difference) and between receivers (double difference) eliminates receiver and satellite clock errors:
18.6.2 Integer Ambiguity Resolution: LAMBDA Method¶
The Least-squares AMBiguity Decorrelation Adjustment (LAMBDA) method (Teunissen, P.J.G. 1995) solves the integer least-squares problem:
The float solution \(\hat{\mathbf{N}}_{\text{float}}\) is obtained from standard LS; LAMBDA decorrelates the ambiguity covariance matrix \(\mathbf{Q}_{\hat{N}}\) using an integer transformation \(\mathbf{Z}\) (\(|\det \mathbf{Z}| = 1\)), then applies sequential conditional least-squares. The fix is validated using the ratio test:
where \(\hat{\mathbf{N}}_1\) is the best and \(\hat{\mathbf{N}}_2\) the second-best integer candidate. A ratio threshold of 3 is commonly used. A confirmed integer fix enables cm-level positioning. RTK baseline limitation: ionospheric decorrelation limits reliable ambiguity resolution to \(\sim 10\)–30 km in standard RTK; Network RTK (NRTK) extends to \(\sim 50\) km using interpolated corrections from a reference station network.
18.7 Precise Point Positioning (PPP)¶
18.7.1 PPP Observation Model¶
PPP uses precise satellite orbits and clocks from an analysis centre (e.g., IGS) instead of broadcast ephemerides, along with dual-frequency measurements to eliminate ionospheric delay:
IGS product accuracy:
-
Final orbits: 2.5 cm (delay: 12–18 days)
-
Rapid orbits: 5 cm (delay: 17 h)
-
Ultra-rapid (predicted): 10–15 cm (delay: real-time)
-
Final clocks: 0.075 ns (5-min sampling)
18.7.2 PPP Convergence and PPP-RTK¶
Traditional PPP requires 20–45 min to converge to centimetre-level precision as the receiver clock, troposphere, and ambiguity parameters are absorbed into the float solution. Integer PPP (PPP-AR) uses satellite phase bias products to recover integer ambiguities, reducing convergence by 30–50% (Ge et al. 2008). PPP-RTK augments PPP with network-derived ionospheric corrections, enabling convergence in \(<5\) min. Commercial implementations: Trimble RTX, TerraStar, GNSS+1, Point One Navigation Atlas.
18.8 GNSS Timing Services¶
GNSS provides time transfer at the sub-nanosecond level, underpinning telecommunications, financial systems, power grid synchronization, and scientific timekeeping.
18.8.1 Atomic Clocks in GNSS Satellites¶
| Satellite | Clock Type | Stability | Drift | Accuracy |
|---|---|---|---|---|
| GPS Block IIF | Rb + Cs | \(10^{-14}\)/day | \(<1\) ns/day | \(<2\) ns |
| GPS Block III | Rb (RAFS) | \(10^{-15}\)/day | \(<1\) ns/day | \(<2\) ns |
| Galileo | PHM (passive H-maser) | \(10^{-15}\)/day | \(<0.45\) ns/day | \(<2\) ns |
| GLONASS-M | Cs + Rb | \(5\times10^{-14}\)/day | 5 ns/day | \(<5\) ns |
| BDS-3 MEO | Rb + PHM | \(10^{-15}\)/day | 1 ns/day | \(<2.5\) ns |
The Galileo Passive Hydrogen Maser (PHM) achieves 1 ps per day frequency stability, the most stable clock ever flown in a navigation satellite (Rochat, P., Droz, F., Wang, Q., Froidevaux, S. 2012). GPS SPS time is specified at \(<30\) ns from UTC(USNO) at 95%.
18.8.2 Two-Way Satellite Time Transfer (TWSTT)¶
TWSTT achieves \(< 1\) ns accuracy by simultaneously transmitting and receiving timing signals through the satellite, cancelling common-path delays. It is the primary method for comparing national atomic timescales contributing to TAI.
18.9 Anti-Jamming and Anti-Spoofing¶
18.9.1 Jamming Vulnerability and Analysis¶
GNSS signals arrive at Earth’s surface at approximately \(-130\) dBm (L1 C/A), some 20 dB below the thermal noise floor of a typical receiver (\(-110\) dBm). The processing gain of spread-spectrum code acquisition is:
The effective jamming range (for a ground jammer of power \(P_J\) and gain \(G_J\) that overloads the receiver) is:
A 1 W omni-directional L-band jammer can disrupt civilian GPS receivers within \(\sim 15\) km radius. Military solutions include:
-
Controlled Radiation Pattern Antenna (CRPA): 7-element array steers nulls toward jammers; \(25\)–\(40\) dB null depth.
-
GPS M-code: BOC(10,5) modulation with 100\(\times\) higher effective power spectral density; encrypted; broadcast from spot-beam regional military satellites for up to 20 dB additional J/S margin.
18.9.2 Spoofing and Authentication¶
GNSS spoofing generates counterfeit signals to mislead the receiver into computing a false position or time. The Galileo OSNMA (Open Service Navigation Message Authentication) (Fernandez-Hernandez, I., et al. 2016) uses TESLA-based delayed key disclosure to authenticate the navigation message:
Keys are disclosed with a delay, preventing replay. Receiver Autonomous Integrity Monitoring (RAIM) detects measurement outliers using a \(\chi^2\) test on the pseudorange residuals. Advanced RAIM (ARAIM) supports Horizontal and Vertical navigation down to LPV-200 precision approach using multi-constellation.
18.10 Multi-Constellation Multi-Frequency (MCMF) Positioning¶
Modern GNSS receivers track all four global constellations plus regional systems, observing 40–80 satellites simultaneously. Benefits include:
-
Improved DOP (more satellites \(\Rightarrow\) better geometry \(\Rightarrow\) lower PDOP).
-
Higher redundancy for integrity (RAIM availability).
-
Improved performance in urban canyons and challenging environments.
-
Faster PPP convergence with triple-frequency observations.
The triple-frequency extra-wide-lane ambiguity (EWL, \(\lambda \approx 5.86\) m) resolves almost instantaneously, bootstrapping narrow-lane (NL, \(\lambda \approx 10.7\) cm) integer resolution within one epoch in ideal conditions.
18.11 Indoor and Urban Positioning¶
18.11.1 Assisted GNSS (A-GNSS)¶
A-GNSS accelerates receiver time-to-first-fix (TTFF) by providing satellite ephemeris, approximate position, and time via a cellular or WiFi network, reducing the need to demodulate the 12.5-minute navigation message (TTFF: cold start 30–50 s \(\to\) A-GNSS \(<2\) s).
18.11.2 GNSS/IMU Sensor Fusion¶
An Inertial Measurement Unit (IMU) provides dead-reckoning between GNSS fixes. The extended Kalman filter (EKF) state vector in a loosely-coupled architecture:
where \(\delta\mathbf{r}\), \(\delta\mathbf{v}\), \(\boldsymbol{\psi}\) are position, velocity, and attitude errors, and \(\mathbf{b}_a\), \(\mathbf{b}_g\) are accelerometer and gyroscope biases. Tightly coupled integration uses raw pseudoranges and Doppler as observations, maintaining position output even with \(<4\) visible satellites.
18.11.3 5G-Based Positioning¶
3GPP Release 16/17 defines radio positioning methods for NR (5G New Radio) (3GPP 2022):
-
DL-TDOA: Downlink Time Difference of Arrival (requires synchronised gNBs, \(\sim 1\)–3 m horizontal).
-
AoD: Angle of Departure using massive MIMO beamforming (\(\sim 1\) m).
-
GNSS hybrid: Combining 5G DL-TDOA with GNSS for \(<0.3\) m (horizontal) in good conditions.
18.12 GNSS Interference and Security Incidents¶
GNSS jamming incidents have proliferated globally: North Korea operates ground-based jammers (Haeju, Kaesong) that have disrupted aviation and maritime navigation in South Korea. GPS jamming in the Baltic Sea region (attributed to Russian electronic warfare) has been reported continuously since 2022, affecting civilian aviation in Finland, Estonia, and Poland. GNSS spoofing of maritime vessels in the Black Sea and Persian Gulf has been documented, where ships reported false positions displaced by tens of kilometres. The Viasat KA-SAT cyber attack (February 2022), which disrupted satellite internet across Europe, demonstrated that ground segment vulnerabilities can cascade to navigation (Viasat Inc. 2022).
18.13 Summary¶
GNSS is a complex system-of-systems spanning space, control, and user segments across four global constellations and multiple regional systems. Key takeaways:
-
The pseudorange observation model links satellite and receiver positions through ranging codes; 4+ measurements resolve the 3D position and receiver clock bias.
-
Error sources (ionosphere, troposphere, multipath) require careful modelling and mitigation; dual-frequency measurements eliminate the dominant ionospheric error.
-
Differential techniques (RTK, PPP) reduce errors to cm or mm level by exploiting carrier-phase ambiguity resolution.
-
SBAS and ARAIM provide integrity assurance for safety-critical aviation applications.
-
GNSS timing underpins global telecommunications, finance, and scientific timekeeping.
-
Jamming and spoofing pose growing threats; signal authentication (OSNMA), antenna nulling (CRPA), and sensor fusion provide resilience.