Mobile Development 5 min read

Understanding GPS Technology and Its Applications in Mobile Location Services

GPS, originally a U.S. military system, now powers smartphones but struggles in urban canyons, tunnels, and elevated roads, so Didi’s Fusion Location Provider blends satellite data with Wi‑Fi fingerprinting, network positioning, dead‑reckoning, and map‑matching to deliver precise, reliable location for ride‑hailing services.

Didi Tech
Didi Tech
Didi Tech
Understanding GPS Technology and Its Applications in Mobile Location Services

GPS is a global positioning system initiated by the U.S. Department of Defense in the 1970s.

It launches 24 satellites in six orbital planes, ensuring that at any point on Earth at least four satellites are visible. By using satellite positions, reception time and time differences, at least three spheres can be drawn in space, and the intersection point gives the receiver's location.

In 2000, U.S. President Bill Clinton announced the cessation of interference with civilian GPS signals, reducing error to about 10 meters, ushering in the civilian era of GPS.

Today, all smartphone CPUs integrate GPS functionality.

GPS‑based positioning solutions are mature; by capturing signals with a baseband chip and an RF chip, a cost‑effective GPS module can be assembled.

In dense urban environments, GPS often suffers severe drift, known as “urban canyon,” where satellite signals reflect off glass surfaces, causing errors of up to hundreds of meters.

Besides urban canyons, tunnels and elevated roads also cause GPS failure. This is critical for ride‑hailing apps like Didi, prompting the development of various solutions.

Didi's Fusion Location Provider (FLP) combines multiple positioning methods—network positioning, road‑network positioning, and vehicle dead reckoning—to compensate for GPS shortcomings in complex scenarios.

Wi‑Fi‑based network positioning (Geo‑Rank) is the most common alternative to GPS. Each Wi‑Fi hotspot has a unique MAC address; the collection of visible hotspots forms a unique fingerprint for a location.

These data allow Didi to transform location problems into machine‑learning ranking problems. Millions of users continuously enrich a database of MAC addresses and GPS locations, while Didi records signal strength, speed, and direction to refine network positioning.

Thus, even when GPS fails, enabling Wi‑Fi lets the device compare with the cloud database to determine location.

Vehicle dead reckoning and map matching further address GPS loss in tunnels and elevated roads. Using onboard inertial sensors (accelerometer, gyroscope), dead reckoning estimates speed and heading; map matching aligns trajectories to road segments.

Every 1–3 seconds, Didi's client uploads a location packet to the real‑time pricing module; after noise reduction and compensation, an accurate trajectory is obtained.

Through these tools, Didi reduces errors in complex scenarios, providing more accurate pickup points and reliable mileage billing.

MobileMachine Learningdata fusionGPSLocation Servicespositioning
Didi Tech
Written by

Didi Tech

Official Didi technology account

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.