Introduction to Satellite Positioning Technology

1. Positioning Technology Overview

Positioning technologies can be broadly classified into two categories based on the type of parameters they measure:

• Absolute Positioning: Provides spatial parameters such as heading, latitude, longitude, and planar coordinates.

• Relative Positioning: Measures motion through 3-axis angular velocity and 3-axis linear acceleration, enabling the estimation of relative movement over time.

To satisfy the navigation requirements of conventional vehicles, manufacturers typically integrate a combination of Global Positioning System (GPS), Inertial Measurement Units (IMU), and wheel speed sensors. These systems jointly support both absolute and relative positioning at the road-level accuracy required for general driving applications.

However, with the evolution of autonomous driving—particularly in Level 2+ systems—the demands for positioning precision increase significantly. Such systems require lane-level positioning, where even minor errors can impact vehicle safety and decision-making. The higher the level of autonomy, the more critical centimeter-level accuracy becomes.

---

1.2 Absolute Positioning

Absolute positioning technologies aim to determine a vehicle’s location in a global coordinate system, typically using satellite-based systems like GNSS. To achieve higher accuracy, Real-Time Kinematic (RTK) technology is often introduced. RTK corrects GNSS data using reference base stations, reducing positioning errors from several meters to just a few centimeters.

When RTK is fused with vehicle-based sensors and high-definition (HD) map matching, it enables precise absolute localization suitable for autonomous navigation.

---

1.3 Relative Positioning

Relative positioning focuses on estimating changes in the vehicle’s pose (position and orientation) over time, without reliance on absolute global coordinates. This is especially useful in GNSS-denied environments such as tunnels, parking structures, or urban canyons.

Advanced relative positioning methods include:

• Visual-Inertial Odometry (VIO): Combines data from cameras and inertial sensors to estimate motion across video frames.

• Laser-Inertial Odometry (LIO): Integrates LiDAR data with inertial measurements, offering higher robustness in textureless or low-light conditions.

These relative positioning systems are often used in tandem with absolute systems to improve robustness and accuracy in complex environments.

Absolute Positioning: In this context, absolute positioning refers to the use of sensors capable of providing three-dimensional coordinate and velocity data within the framework of the International Terrestrial Reference Frame (ITRF). For the sake of clarity and focus, this document does not differentiate between single-point positioning and differential positioning methods such as RTK or PPP.

Relative Positioning: Relative positioning involves sensors that do not yield absolute spatial coordinates but can infer 6 Degrees of Freedom (6-DoF) pose changes—namely, translation along and rotation around the x, y, and z axes—over time. These measurements allow the system to estimate motion relative to a previous state, making them critical for continuous localization in environments where absolute signals may be unreliable or unavailable.

---

2. Satellite Positioning

Satellite-based positioning technologies encompass a variety of systems, including Global Navigation Satellite Systems (GNSS), Differential Global Positioning System (DGPS), Real-Time Kinematic (RTK), and Precise Point Positioning (PPP). These systems differ in architecture, accuracy, and use case, but all fundamentally rely on signals from orbiting satellites to determine position.

2.1 Global Navigation Satellite System (GNSS)

GNSS is a general term that includes all satellite navigation systems—both global and regional—as well as satellite-based augmentation systems (SBAS). These systems provide users with three-dimensional positioning, velocity, and time (PVT) information anywhere on Earth or in near-Earth space, regardless of atmospheric or environmental conditions. Examples of GNSS include GPS (United States), GLONASS (Russia), Galileo (European Union), and BeiDou (China).

Nation	Global Positioning System	Regional Positioning System	Augmented Positioning System
USA	GPS	N/A	WAAS (Wide Area Augmentation System)
Russia	GLONASS	N/A	SDCM (Satellite-based Differential Correction Message)
China	COMPASS	N/A	BDSBAS (BeiDou Satellite-based Augmentation System)
Europe	Galileo	N/A	EGNOS (European Geostationary Navigation Overlay Service)
Japan	N/A	QZSS	MSAS (MTSAT Satellite-based Augmentation System)
India	N/A	IRNSS	GAGAN (GPS Aided GEO Augmented Navigation)

2.2 GNSS Positioning Principles

GNSS positioning is based on pseudo-range measurements, which are susceptible to a variety of error sources, including:

• Ephemeris errors (inaccuracies in the broadcast satellite orbit parameters)

• Satellite clock errors

• Atmospheric disturbances, including ionospheric and tropospheric delays

These measurements involve four unknowns: the three-dimensional coordinates of the receiver (X, Y, Z) and the receiver clock bias. Consequently, a minimum of four satellites is required to compute a valid position solution.

---

2.3 GNSS Signal Composition

GNSS signals consist of the following three primary components:

1. Carrier Wave

   • The GPS system operates primarily on two frequency bands:

     • L1: 1575.42 MHz

     • L2: 1227.60 MHz

   • These carrier frequencies serve as the foundation for code modulation and are also used in advanced techniques such as carrier-phase positioning.

2. Pseudo-Random Noise (PRN) Code

   • Encoded signal used to determine the time of flight from satellite to receiver.

   • Two main types:

     • C/A (Coarse/Acquisition) Code:

       • Frequency: 1.023 MHz

       • Length: 1023 bits

       • Designed for civilian applications

     • P (Precision) Code:

       • Frequency: 10.23 MHz

       • Primarily used for military applications

3. Navigation Message

   • Transmitted at 50 bits per second

   • Contains essential satellite and system data, including:

     • Satellite health status

     • Ephemeris and almanac data

     • Satellite clock correction terms

     • Ionospheric model parameters

The theoretical lower bound for GNSS single-point positioning accuracy using the C/A code is approximately 10 meters Circular Error Probable (CEP). This limitation arises from multiple error sources, including satellite orbit and clock inaccuracies, signal multipath, receiver noise, and atmospheric effects such as ionospheric and tropospheric delays.

3. GNSS Differential Augmentation Systems

To enhance positioning accuracy, differential augmentation systems such as Ground-Based Augmentation System (GBAS) and Satellite-Based Augmentation System (SBAS) are widely employed.

• GBAS: This system deploys multiple reference stations with precisely surveyed coordinates to monitor satellite signals. By calculating and broadcasting correction data for errors including ephemeris inaccuracies, satellite clock errors, and ionospheric disturbances, GBAS significantly improves positional precision for users within its coverage area.

• SBAS: Similar to GBAS, SBAS uses a network of ground reference stations to collect satellite data and compute error corrections. These corrections are then transmitted to users via geostationary satellites, providing wide-area augmentation that enhances GNSS positioning accuracy across broader regions.

GABS https://www.faa.gov/about/office_org/headquarters_offices/ato/service_units/techops/navservices/gnss/laas/howitworks

SBAS https://www.eoportal.org/other-space-activities/sbas

4. Real-Time Kinematic (RTK)

RTK is an advanced carrier-phase differential positioning method that utilizes real-time carrier phase measurements from a reference station with known coordinates. These measurements allow for the correction of common GNSS errors, which are then broadcast to mobile receivers to achieve highly accurate positioning.

RTK typically achieves centimeter-level accuracy—on the order of 1 cm plus 1 part per million (ppm) of the baseline distance. This superior accuracy is attributed to the high-frequency L1 carrier signal used in RTK, which enables much finer range measurements compared to the lower-frequency C/A code.

RTK https://pointonenav.com/news/rtk-survey/

5. Precise Point Positioning (PPP)

While Real-Time Kinematic (RTK) technology provides high accuracy positioning, its performance tends to degrade as the distance between the rover and the base station increases. Precise Point Positioning (PPP) addresses this limitation by leveraging a single dual-frequency GNSS receiver in conjunction with precise satellite ephemeris and clock correction products provided by the International GNSS Service (IGS).

PPP enables accurate absolute positioning without the need for a nearby reference station, making it especially useful in remote or expansive areas where establishing ground-based infrastructure is impractical. This technique achieves decimeter to centimeter-level accuracy through advanced processing algorithms that mitigate satellite orbit and clock errors.

PPP https://www.e-education.psu.edu/geog862/node/1841

	GNSS	GBAS	SBAS	RTK	PPP
Principle	Single-point Positioning	Differential positioning C/A code	Differential positioning C/A code	Differential positioning Carrier phase	Single-point Positioning Carrier phase
Accuracy	Meter-level	Decimeter-level	Decimeter-level	Centimeter-level	Decimeter-level
Received Content	GNSS Signals (C/A code)	GNSS Signals (C/A code) The total error correction from base station	GNSS Signals (C/A code) The total error correction from base station	GNSS Signals (Carrier phase) The total error correction from base station	GNSS Signals (Carrier phase) The precise ephemeris and clock error products from IGS
Applicable Scenarios	Global	Within a radius of 40 km from the reference station	Global	Within a radius of 40 km from the reference station	Global
Initialization Time	Second-level (No need to set up base stations)	Second-level (After the base station network is established)	Second-level (After the base station network is established)	Second-level (After the base station network is established)	30 min (No need to set up base stations)
Disadvantages	Low positioning accuracy	Requires the deployment reference stations	Incomplete coverage in high latitude areas	Requires the deployment reference stations	Long initialization time