Everything you want to know about satellite positioning


Introduction:This article will briefly introduce the principle and application of satellite positioning, so that everyone can have a better understanding of Beidou and satellite positioning.

5g and Beidou are the most important weapons of the country. As a satellite positioning system, Beidou is now in a leading position in the world, and has penetrated into all aspects of our work and life. This article will briefly introduce the principle and application of satellite positioning, so that everyone can have a better understanding of Beidou and satellite positioning.

Principle of satellite positioning

The English of satellite positioning system is global navigation satellite system (GNSS). Although it is directly translated as navigation satellite system, its real capability is positioning. After positioning, navigation becomes relatively simple. The principle of satellite positioning is to use the satellite broadcast time signal. When the equipment receives it, it can calculate the signal transmission time according to the signal transmission time and local time, and then obtain the satellite equipment distance in combination with the speed of light.

Everything you want to know about satellite positioning

With the signals from multiple satellites, we can list a set of equations and solve four unknowns: the three-dimensional coordinates of the equipment x/y/z, and the time difference between the local time and the GNSS system.

Where represents the three-dimensional coordinates of satellite J, which can be obtained through the calculation of satellite ephemeris.

Everything you want to know about satellite positioning

Ephemeris is a set of parameters that describe the orbit of a satellite. The orbit of a satellite is an ellipse. Through several parameters and time, the exact position of a satellite can be uniquely determined.

There are two ways to obtain the ephemeris. One is direct satellite broadcasting. The advantage of this method is that the positioning process does not rely on any input other than the satellite signal. Even if there is no network, the positioning can be successful, but the problem is that the satellite link bandwidth is very small. It takes about 30 seconds to download the complete ephemeris. The positioning process of early mobile phones and some on-board devices is very slow, which is why.

Another way is to broadcast through the Internet. This way is called a-gnss, and the specific transmission protocol is SUPL (secure user plane location). This kind of data is generally not disclosed to the application layer. On the mobile phone, the operating system will regularly request SUPL data at the bottom layer, and then inject the obtained ephemeris into the GNSS chip. With a-gnss, the device can get location in seconds without any waiting process. At present, all mobile phones support this method. The service providers of a-gnss are mainly communication operators and some location-based service providers, such as Google, Qianxun, etc.

The satellite continuously broadcasts signals to the ground, which mainly include the following information:

  • Satellite number. It is used to find the satellite orbit from the ephemeris, and then obtain the current satellite position combined with the time stamp
  • Current timestamp. It is used to obtain the satellite position and calculate the pseudo range on the other hand. Pseudo range is (local time – signal transmission time) * the speed of light. It is called pseudo range because the local time is not synchronized with the satellite time, so this distance is not the real device satellite distance.
  • Ephemeris data. Used to calculate satellite position.

Like all other communication technologies, this information is also sent in the form of messages. Take GPS as an example, the satellite will send a packet every 6 seconds, and this packet will be decomposed into data bits CA code sequence carrier waveform, which will be transmitted to the ground through the antenna. The ground equipment continuously locks the satellite. During the calculation, calculate the timestamp of each satellite at the current time (add the latest received timestamp to the message offset), and then calculate the position.

Everything you want to know about satellite positioning

The carrier frequency is about 1.5g, and the wavelength is about 20cm, which is slightly longer than the wavelength of mobile communication, so the signal penetration is relatively good (the longer the wavelength, the easier it is to bypass obstacles), and it can penetrate relatively thin walls or roofs, so in some cases, it can be positioned even if the sky cannot be seen directly. However, the satellite signal is from top to bottom, and it is difficult to pass through multi-storey buildings indoors.

Another feature of satellite positioning is that it can calculate the velocity, which is based on the principle of Doppler frequency shift (the same principle as the velocimeter used by traffic police). When there is relative movement between the signal source and the receiving equipment, the frequency of the received signal will change.

Everything you want to know about satellite positioning

There is the following formula between frequency variation and relative speed:

Everything you want to know about satellite positioning

Among them, the left side of the formula is the frequency difference and wavelength, V is the equipment motion speed (vector), VJ is the satellite motion speed (vector), 1J is the projection direction of the satellite, and dt’is the frequency drift speed of the local equipment. As long as the frequency difference of 4 stars is measured, the motion speed of the local equipment can be solved (independent of the equipment attitude).

In addition to positioning and speed measurement, positioning satellites can also complete global time service (local clock difference is obtained in the process of calculation), which is also the lowest cost high-precision time service method at present, which is more accurate than the clocks of most equipment.

Generally speaking, the accuracy of pseudo range measurement is not as high as that of frequency measurement (the pseudo range positioning accuracy is about 10 meters, while the Doppler speed fixing accuracy can reach within 0.2 meters / second, and the timing accuracy is 20ns), because pseudo range measurement is vulnerable to a variety of path errors (described later), and the interference factors of frequency measurement are much less.

Development history of satellite positioning

The earliest satellite positioning system was the meridian system developed by the United States in the 1960s, followed by the GPS positioning system developed in the 1970s. The current GPS system consists of 24 satellites. In addition to GPS, many countries in the world have also developed their own satellite positioning systems, mainly China’s Beidou system, the EU’s Galileo system, and Russia’s glonaz system. In addition, Japan and India are developing regional positioning systems.

Everything you want to know about satellite positioning

In addition to the satellites in the sky, each positioning system also needs the ground station to monitor the operation of the satellite, including health, orbital parameters (injected into the satellite after calculation to realize global broadcasting), signal quality, etc. in addition, it also needs to control the satellite.

The technologies used by various satellite positioning systems are similar, most of them adopt medium orbit satellites (MEO, satellite altitude of 20000 km), and a few adopt geosynchronous orbit (GEO, satellite altitude of 40000 km) and earth tilt synchronous orbit (IGSO). At the same time, most of the signal broadcasting uses CDMA technology to transmit signals from multiple satellites on the same frequency. In order to enable ground equipment to better receive signals from tens of thousands of kilometers away, the data rate of the signals is relatively low. For example, the data transmission rate of GPS L1 band is only 50 bytes /s. According to Shannon’s theorem, c=b*log2 (1+s/n). When the frequency bandwidth B is fixed, with the reduction of transmission rate C, the receiver can also solve the correct signal when the signal-to-noise ratio (s/n) is relatively low, which is conducive to the continuous locking of satellite signals.

Compared with other positioning systems, Beidou has the following characteristics:

  • The Asia Pacific region is well covered. Beidou system is composed of three geostationary satellites, three earth tilt orbit synchronous satellites and 24 medium orbit satellites. Compared with GPS, Beidou has 6 satellites continuously covering the Asia Pacific region, which greatly improves the number of visible satellites in the Asia Pacific region. On the one hand, it can improve the success rate of positioning, on the other hand, it can also improve the accuracy (improve GDOP and reduce errors).
  • Beidou’s synchronous satellite can be used for communication. The ground equipment can send short messages to the satellite (only GEO Satellite supports short messages) and then forward them to the target terminal. This kind of communication is free, but it needs special antennas and equipment (it needs to send signals to 40000 kilometers away, and ordinary mobile phones certainly can’t).

Everything you want to know about satellite positioning

When the signals of multiple satellite positioning systems are received at the same time, all satellites can participate in the solution together (for each additional system, only a new parameter needs to be added, that is, the time difference between the system and the GPS system), so that the positioning accuracy can be improved. At present, the constellation or satellite involved in positioning cannot be selected on the mobile phone, so we cannot specify to use only Beidou or GPS positioning.

We compared the proportion of different systems used in GNSS positioning on the mobile terminal. It can be seen that GPs and glonaz are also used in the highest proportion due to their early development and high penetration rate on the mobile chip side, followed by Beidou.

Everything you want to know about satellite positioning

According to the number of satellites involved in positioning, Beidou ranks second, second only to GPS.

Everything you want to know about satellite positioning

Because the technology of each system is similar, its positioning accuracy is also similar, and Beidou is no exception. The horizontal positioning error is generally within 10 meters. The accuracy of vertical positioning is generally poor, mainly because the satellites are distributed on one side of the equipment, and the error in the vertical direction is difficult to correct.

Composition of satellite positioning receiver

Everything you want to know about satellite positioning

The schematic diagram of the satellite positioning receiver is shown in the figure above. The main modules include:

  1. antenna

Used to receive satellite signals. Because the satellite signal is weak, the larger the antenna is, the better, but because the receiver needs to move, the size of the antenna is restricted. The main functions of antennas are to amplify signals and suppress multipath. The main types are as follows

Everything you want to know about satellite positioning

On the left is a common antenna, with a ceramic antenna inside and a magnet outside, which can be adsorbed on the roof; In the middle is a professional antenna with a choke next to it, which can suppress the signals reflected from the surrounding and the ground, and only receive the signals from the zenith direction. This antenna has the best effect, and is generally used for professional research and high-precision mapping; On the right is the mobile phone antenna, which is only a few centimeters long, and the effect is the worst.

The electromagnetic wave of satellite signal is circularly polarized (it fluctuates on a plane perpendicular to the propagation direction when propagating). Therefore, the receiving effect of circularly polarized antenna (such as planar ceramic antenna) is the best. However, the size of the antenna on the mobile phone is too small, so the linear polarization antenna can only be used, which greatly reduces the signal acquisition ability. Coupled with the lack of signal shielding (choke), it is very vulnerable to multipath effect and other signal interference.

  1. RF front end

This module mainly performs down conversion, power amplification and filtering of the original signal to extract the really useful signal for decoding and processing.

  1. Baseband processing

This module decodes the satellite signal and obtains the satellite message. The signal of each satellite needs a separate channel for processing. If there are 100 satellites and 2 frequency bands, it may need 200 channels to effectively process this information. The more channels, the richer satellite observations can be obtained, and the higher the positioning accuracy.

The decoding process is divided into three steps: search lock analysis. First, the pseudo code of each satellite is generated, and then the signal is autocorrelated. When the correlation reaches a certain degree, the satellite can be locked, and then the code locking, bit synchronization, frame synchronization, and finally the message is extracted. This process should continue. Because of the Doppler effect, the frequency of the signal will constantly change, so the locally generated pseudo-random code also needs to constantly change the frequency to adapt to the changes of the satellite. Once the lock is lost, the signal will be lost and the location will be impossible.

  1. PVT solution

PVT includes position, velocity and time. This step is the real positioning step. It uses the message obtained by baseband decoding to extract the timestamp, ephemeris and other information, substitute it into the formula for calculation, and then output the calculation result to the application program.

Source of positioning error and accuracy improvement

Although satellite positioning has been very accurate, it still cannot meet the needs in some scenarios. For example, the positioning point is a certain distance from the vehicle when taking a taxi, it is difficult to distinguish the direction when walking, and it will even be positioned across the road when stationary, the total number of positioning points will float around when stationary, and the positioning points will float around when indoors. This needs to be explained from the process of satellite signal transmission, transmission and reception.

Everything you want to know about satellite positioning

Satellite signals need to pass through the atmosphere from launch to receiving by equipment. Among them, the ionosphere of the atmosphere is thousands of kilometers thick. This part of the atmosphere is very thin, but there are a large number of ionized electrons, which will slow down the electromagnetic wave, resulting in delay. In the troposphere, there will also be a certain delay. Near the earth’s surface, due to the influence of various buildings, mountains and water surfaces, satellite signals may be reflected or refracted (multipath effect), resulting in delay.

There are also many system related errors on the transmitting side and receiving side of the satellite signal, such as clock deviation, processing delay, etc. these delays plus the transmission delay make the transmission time of the satellite signal not exactly equal to the physical distance / speed of light. On the other hand, the ephemeris of the satellite also has errors, and there is a deviation between the satellite position and the real position, which eventually leads to the deviation of the positioning results.

To improve the positioning accuracy, we need to find ways to eliminate these errors. There are mainly the following schemes.

Dual band GNSS

Electromagnetic waves of different frequencies will have different delays when passing through the ionosphere. It is found that the linear combination of the observed values of two or more frequencies can eliminate the ionospheric error and improve the accuracy. This is the principle of dual frequency GNSS positioning. Xiaomi 8 is the first mobile phone in the industry that supports dual band GNSS positioning. Subsequent major manufacturers have followed up, and some high-end mobile phones use dual band positioning. After eliminating the ionospheric error, the positioning accuracy can be improved to less than 5 meters.

Foundation / Satellite Based Reinforcement

Ephemeris error, satellite clock error, and even ionospheric and tropospheric errors can be observed or modeled. Once the real-time error value is calculated, it can be broadcast through a separate channel. The receiving equipment can improve the positioning accuracy by using these correction items in the positioning process. Generally, there are two channels for broadcasting. One is to broadcast directly through satellite, called SBAS (satellite based augmentation system), which has the advantage of wide coverage, but the equipment needs to add additional signal receiving channels; The other is foundation enhancement, such as through the Internet, which requires the equipment to have networking capabilities.

These enhancement methods are limited for accuracy improvement, or there are many error terms that cannot be eliminated, such as ionospheric error.

High precision positioning differential positioning (RTK)

RTK is the abbreviation of real – time kinematic, which is a kind of differential positioning. The principle is to use a reference station to provide the reference observation value, and then use the observation value of the equipment to make a difference with the observation value of the reference station. After the difference, the ephemeris error, satellite clock error and ionospheric error can be eliminated, and then the clock error of the equipment can be further eliminated after the inter satellite difference. Finally, the relative coordinates of the equipment relative to the reference station can be calculated. If the position of the reference station is known, the accurate absolute coordinates can be completed, The accuracy can reach centimeter level or even millimeter level.

Another reason why RTK can improve accuracy is the introduction of carrier phase observation. Compared with pseudo range observation, the error of carrier phase observation is smaller.

Using RTK requires a reference station within 20km nearby (the distance is too far, the ionospheric error is different, and the difference cannot completely eliminate the error). At the same time, it needs to continuously obtain the observation data of the reference station (usually transmitted through the Internet, using RTCM protocol). Therefore, compared with ordinary positioning, RTK Positioning costs more, but for some scenes that require high accuracy, such as lane level positioning, automatic driving, etc, Is essential.

RTK services are generally provided by professional service providers, such as Chihiro and Liufen. These service providers have deployed thousands of benchmark stations nationwide to continuously broadcast data to subscribers.

High precision positioning – precision single point positioning (PPP)

RTK needs to deploy dense reference stations. Is there any way to not rely on reference stations? PPP (precise point positioning) is a method. Its principle is to accurately model each error and finally solve the accurate distance between the satellite and the equipment. In order to determine the accurate error, PPP positioning requires continuous iteration of internal parameters. Moreover, the error of some satellites can be reflected only after the satellite position changes, so PPP needs a long convergence time, which generally takes 30 minutes to converge to the ideal accuracy. How to converge faster is a research hotspot in the academic community at present.

Combined positioning

One of the biggest problems of satellite positioning is how to locate after losing satellite signals, which needs to be supplemented by other positioning methods. Integrated positioning uses satellite signals and other positioning technologies, such as inertial navigation, to complete positioning, and the two cooperate with each other. The simplest example is that satellite positioning has a highest frequency, usually 10Hz at most. Between two positioning, inertial navigation can be used to calculate the position and obtain higher frequency position output. The most important role of integrated navigation is to improve the accuracy. For example, using Kalman filtering method, using inertial navigation to calculate the estimated position, using satellite positioning to provide observation, and correcting the calculated position, which can make the positioning result smoother, and can filter or reduce the weight of abnormal satellite observation.

Satellite positioning on mobile phones

Before the emergence of mobile Internet, satellite positioning terminal was a very professional field. Only surveying and mapping, military and other fields would apply this technology. Positioning needed to use special receivers, such as Trimble, ublox and so on. With the integration of smart phones with satellite positioning chips, the application of satellite positioning has exploded, and the number of terminals has suddenly increased to several billion, resulting in massive location data.

There are still big differences between satellite positioning on mobile phones and professional receivers, mainly reflected in:

  • The size of the mobile phone is limited, the antenna is relatively small, and the ability to capture, lock and denoise the original signal is relatively poor, resulting in the quality of the received signal is naturally inferior to the professional receiver.
  • The cost of chips on mobile phones is relatively low, and the number of supported channels is relatively limited. The number of satellites and systems that can be decoded at one time is relatively small, mainly single frequency, a few dual frequency, and no triple frequency.
  • Mobile phones have high requirements for power consumption and performance overhead, and cannot spend a lot of resources on positioning. The complexity of the solution algorithm is relatively low, the effect is relatively limited, and the accuracy is relatively poor.


The positioning capability of Apple mobile phone is completely closed. Only the positioning results are revealed to the outside, and the outside basically cannot get any original observation related to positioning, such as the number and type of satellites. The good news is that iphone12 is finally supporting Beidou. From Apple’s API, the outside world can’t even tell whether the positioning results come from satellite positioning or network positioning (at present, it can only be judged by the symbol of speed, but Apple has no commitment to this). Therefore, based on Apple mobile phones, we can hardly make optimization. The positioning points of Gaode map on Apple mobile phones are directly provided by the bottom of IOS.

Android phone

Android phones are more open than apple phones, and provide a series of APIs in terms of location capabilities:

  • Satellite positioning results or network positioning results can be obtained separately, or both kinds of positioning can be carried out at the same time.
  • It provides the result data in NMEA format (a standardized expression of satellite positioning results), which can obtain the ID, type, signal strength, xdop and other fine-grained error descriptions of each satellite.
  • Gnsstatus is provided to describe the status of each satellite, which is more comprehensive than NMEA.
  • Gnssmeasurement is provided to describe the original observation, including pseudo range measurement value, carrier phase measurement value, satellite locking state, etc.
  • Gnssclock is provided to describe the status of the local clock.
  • It provides the most original decoded message revealed by gnssnavigation.

With this information, you can see the current satellite status in real time through some apps, such as androits GPS test, gpstest, etc

Everything you want to know about satellite positioning

In addition, we can also carry out the soft solution of satellite positioning, and modify or even replace the satellite positioning results. We mainly try to solve two kinds of problems:

  • Inaccurate positioning: judge the quality of the satellite positioning results, identify the major errors, optimize the accuracy, or optimize the accuracy radius, so that the downstream can handle the positioning points differently.

Everything you want to know about satellite positioning

The main reason for the inaccurate positioning is that the satellite signal contains errors, and the most serious and difficult to suppress is the influence caused by multipath.

Another kind of inaccurate positioning problem is that the system forges other positioning results into satellite positioning results. For example, fake network points as satellite positioning points.

  • Unable to locate: when the system does not output the positioning results, try to perform software solution.

The main reason for the inability to locate is the poor signal receiving conditions, such as indoor shelter, overhead shelter, and high-rise shelter. Unable to locate in open areas, it is usually a device bug, which can generally be solved after restarting the device.

Future prospects of satellite positioning

With the continuous growth of mobile users and the wide spread of the Internet of things, satellite positioning technology will continue to develop rapidly.

On the satellite side, Leo positioning satellites (hundreds of kilometers from the ground) will appear. The traditional positioning satellites usually have a high altitude and operate in the medium orbit because they cover a large geographical range. With the revolution of rocket launch technology, the cost of satellite launch has fallen sharply, and it is possible to launch large quantities of low-cost satellites into space. For example, SpaceX has launched thousands of “star chain” satellites.

LEO satellite positioning has several advantages:

  • The distance is close, the signal is stronger, and the satellite signal received by the equipment side is better.
  • More data can be transmitted, such as various correction data.
  • Located at the bottom of the ionosphere, the ionospheric error is small.
  • With the rapid change of satellite elevation, PPP positioning can converge faster.

On the device side, high-precision positioning will be popularized on a large scale. Huawei P40 is the first smart phone supporting RTK, which can achieve an accuracy of 0.5 meters. Qualcomm is also about to release RTK enabled mobile chips. In the first half of 2021, more RTK enabled smartphones will be available.

On the application side, high-precision positioning application scenarios will continue to emerge. Some typical application scenarios now include:

  • Traditional surveying and mapping
  • Precision agriculture, mechanized automatic planting and harvesting
  • Lane level navigation and autonomous driving
  • Precise parking of shared bicycles
  • UAV navigation

Original link
This article is the original content of Alibaba cloud and cannot be reproduced without permission.