Have you ever used your phone's navigation system to find an address or landmark, only to find that the location you're looking for is across the street from the phone's display? In fact, when we understand how the mobile phone's location function works, this will not be surprising.
How targeting works (and why it sometimes doesn't)
Global Navigation Satellite System (GNSS) receivers, whether in smartphones, cars, or other devices, calculate their position from signals from at least four active GNSS and regional satellite systems. [1]
The receiver calculates the distance or range from the satellite to the receiver as the time difference between when the satellite sent it on air and when the signal was received. When a GNSS device receives signals from many satellites, the GNSS device knows where each satellite is, and how far it is from them. The more satellite signals GNSS receives, the more accurate the positioning will be.
GNSS global navigation satellite systems include Global Positioning System (GPS), Galileo Positioning System (Galileo), Beidou Satellite Navigation System (BeiDou) and Russia's Global Navigation Satellite System (GLONASS). Among them, each satellite of GPS, Galileo and Beidou transmits its signal through different frequency bands (up to three). This means that older receivers that only support a single band for GPS and GLONASS are not as accurate as newer devices that support four satellite systems and dual-band systems.
Because GNSS satellites orbit 24,000 kilometers (14,913 miles) above Earth, the signals reaching the ground are too weak to pass through rooftops, dense woods or tunnels. In addition, the time it takes for a signal to travel through the ionosphere and troposphere also affects it.
In dense cities, high-rise buildings can also block the signals of some satellites, reducing the number of signals used to calculate the position. In addition, high-rise buildings can create a canyon effect, causing signal reflections and multiplication between the satellite and the receiver, and these "echoes" will affect the accuracy of the distance.
The quality of GNSS in a smartphone depends on the GNSS receiver (chipset), positioning engine (software), antenna quality and hardware integration. In addition to pure GNSS positioning, current smartphones use inertial sensors (such as gyroscopes, accelerometers, and barometers) as well as network positioning (relying on cellular networks and Wi-Fi). These different technologies can compensate for the lack of GNSS satellite signals.
Of course, mobile phones didn't have built-in navigation systems to begin with. The first mobile phone to combine GPS functionality with built-in maps is from Finnish mobile phone manufacturer Benefon - the Benefon ESC! It came out in 1999, the main sales market is in Europe, and laid the foundation for the mobile phone equipped with GPS. [2]
In 2011, the MTS 945 GLONASS, built by Qualcomm and ZTE for Mobile TeleSystems, was the world's first GLONASS-compatible smartphone. [3]
The Mi 8, launched in 2018, is the first dual-frequency GNSS smartphone. It is equipped with a Broadcom BCM47755 chip, which can provide decimeter-level accuracy for mobile positioning services and vehicle navigation. [4]
These days, all but the cheapest smartphones come with some sort of navigation system. While DXOMARK's smartphone testing guidelines do not include evaluating the accuracy of a phone's GPS positioning, GPS navigation is one of the use cases evaluated in our battery tests. Therefore, not long ago, DXOMARK cooperated with Geoflex, a French company that is quite professional in geolocation and precise positioning, to evaluate the accuracy of smartphone geolocation.
Test Methods
DXOMARK collected a variety of mobile phones from different brands and price points to understand the overall situation of GPS positioning performance of smartphones.
In order to obtain a high-precision reference trajectory, DXOMARK used the Geoflex test platform and installed it on the top of the car, as shown in the following figure:
Geoflex test platform
The test platform consists of the following parts:
- The most accurate multi-satellite system (GPS, Galileo, GLONASS, BeiDou) and tri-frequency GNSS receiver on the market, capable of processing the latest signals from the European Galileo constellation (E6); on top of this hardware platform, the Geoflex algorithm calculates accurate Positioning point, the accuracy can reach below 4 centimeters. In order to compare with the position provided by the smartphone under test, it is necessary to obtain a good reference trajectory.
- High quality GNSS antenna;
- GNSS is combined with inertial systems to improve reference trajectories when GNSS signal quality is poor or non-existent.
We installed the test platform in the center of the car so that we can compare the trajectory of all smartphones at the same point on the central axis and the horizontal axis. We use the data recorded in the setup as a control or reference to compare the GPS performance of the smartphones. We installed the smartphone on the board in the front right of the car, as shown in the picture below:
Dashboard setup for testing.
We then set out on a three-hour drive through the greater Paris region with reference systems and smartphones. Our itineraries included urban areas, "urban dense" areas (tall buildings close together), open areas, and sheltered areas like tunnels, forests, and underpasses. Except for open areas, every other area is a challenge for the GPS system.
Our itineraries include a variety of environments: including urban areas, "urban dense" areas (where tall buildings are close together), open areas, and sheltered areas such as tunnels, forests, and underpasses
We use GNSS Logger software to extract the data we collect, which includes:
- Positioning solutions in NMEA format [5] generated by each smartphone
- GNSS observation archive containing satellite data
Additionally, we analyze each trajectory to produce:
results and details
DXOMARK focuses on the positioning error on the horizontal plane and measures the RMS (root mean square) error without considering the height error.
The graph above illustrates the difference we observed:
- The dark green area represents very good phones with an average error of less than 5 meters of horizontal error (RMS).
- The light green part shows the mobile phones with an error of more than 5 meters to nearly 8 meters (RMS).
- Finally, the red part shows the mobile phone that assumes a linear displacement and continues to locate the position even though the tunnel is curved. The phones had a horizontal error of 450 meters at the exit of the tunnel, which is more than 20 meters away from the average.
Unsurprisingly, high-end smartphones with faster processors, better antennas, more constellations, and dual-band systems can deliver better GPS performance than more affordable phones. Since GNSS and SNR (signal-to-noise ratio) propagation conditions are very good, all smartphones perform well in open areas. However, in more difficult environmental conditions, especially when passing through tunnels, the performance of each mobile phone varies significantly. The diagram below shows the comparison of the positioning results of three smartphones with the reference positioning results (red line): mobile phone 5 (best performing, yellow), mobile phone 10 (medium performing, purple) and mobile phone 15 (in The worst performer in our tests, blue).
Reference (red); Mobile No. 5 (yellow); Mobile No. 10 (purple); Mobile No. 15 (blue)
Mobile phone No. 10 stopped positioning in the tunnel (acceleration sensor has reached the accuracy limit), while mobile phone No. 15 continued to show a straight-going positioning without a reference trajectory (probably due to the single-axis accelerometer).
In contrast, mobile phone 5 has a 3D gyro sensor and accelerometer, which can follow the arc of the tunnel, which is different from the trajectory without GNSS reception. In fact, mobile phone 5 also showed the closest positioning result to the reference trajectory in dense urban areas, because its inertial sensor was well positioned between buildings.
Reference (red); Mobile No. 5 (yellow); Mobile No. 10 (purple); Mobile No. 15 (blue)
Where to go from here?
The advent of quad-constellation and dual-band navigation satellite technology means that earlier satellite navigation systems have come a long way; even with newer GNSS receivers (including those capable of receiving tri-band signals) on more affordable smartphones , will also be more common. But, as the next generation of devices incorporates more advanced navigation technologies and the need for more precise positioning grows, what other advances can we expect to see?
In addition to smartphones, we also foresee that many embedded electronics and devices such as 360° cameras will widely use GNSS technology. We believe that GNSS correction services will help manufacturers and end users augment satellite signals and inertial sensor data so that positioning accuracy can be measured in centimeters rather than meters or greater distance units. Such accuracy will have a huge impact on human mobility as well as a wide variety of services such as drone delivery.
When it comes to the audio, battery, camera or screen experience of a smartphone, the hardware is only one part, and the software is the core part. We can see that phones with the same hardware don't show the same accuracy, it's all about software tuning.
The latest processors are also expected to improve accuracy. The Snapdragon 8 gen 2 has lane-level accuracy and can pinpoint location to within +/- 1.5 meters, which is a device we'd love to test.
DXOMARK expects the next generation of positioning technology to take us further!
[1] Global Navigation Satellite Systems: GPS (Global Positioning System, developed by the United States); Beidou (developed by China); Galileo (developed by the European Union); GLONASS (developed by Russia). Regional satellite systems: QZSS (Quasi-Zenith Satellite System, developed by Japan); IRNSS (Indian Regional Navigation Satellite System, developed by India)
[2]Mobile Phone Museum | Mobile Phone Museum
[3] Wikipedia, "MTS 945" https://en.wikipedia.org/wiki/MTS_945
[4] World’s first dual-frequency GNSS smartphone hits the market | European GNSS Service Centre
[5] NMEA, defined by the National Marine Electronics Association, is a standard data format followed by all GPS manufacturers, much like the standard for digital calculator characters in the calculator industry. The purpose of NMEA is to allow device users to pair hardware and software.
[6] The Earth-Centered Earth-Fixed (ECEF) coordinate system is also known as the "conventional Earth" coordinate system.
[7] KML stands for "Keyhole Markup Language," a file format used by Google developers to display geographic data.
Founded in 2012 , Geoflex is a French company that provides comprehensive hyper-geolocation services to many different industries, including transportation, smartphones, robotics, agricultural machinery and more. The company works with leading global players (private and public) who need precise positioning to enhance and enhance their vital business processes. Smart
Geoflex - Hypergeolocation everywhere