Qualcomm Patents a Cross-Check System That Catches False Location Readings
GPS can confidently report the wrong location, and the system has no idea. Qualcomm's new patent tackles that uncomfortable truth by making two satellite receivers cross-examine each other.
How Qualcomm's dual-receiver GPS safety check works
Every time your car's navigation system locks onto a GPS signal, it reports back a position along with a confidence level. The problem is that a GPS receiver can be highly confident about a position that is simply wrong, thanks to signal reflections off buildings or other interference. Your device has no built-in way to know it has been fooled.
Qualcomm's patent describes a system that uses two GPS receivers at the same time. Because their physical separation is fixed and known in advance, the system can compare what each receiver reports. If both receivers say they're confident in their fix, but their reported positions don't match what the known distance between them predicts, the system raises a red flag called a wrong fix indication.
Think of it like two witnesses who both claim they saw the same event clearly, but whose accounts contradict each other. When both are confident and wrong, the disagreement exposes the error neither could catch alone.
… determine a horizontal offset value based on the determined antenna baseline vector and a difference between the determined first position estimate and the determined second position estimate; and generate a wrong fix indication …
Translation: It compares the distance between the two antennas against the calculated positions to spot errors.
How the two receivers catch each other's bad position locks
The system works by exploiting a fixed physical relationship between two antennas. Before anything else, the device records the antenna baseline vector, which is just a precise description of how far apart the two antennas are and in what direction relative to each other.
At the same moment, each antenna feeds its satellite signals to its own GNSS receiver (a GNSS receiver is any device that interprets signals from satellite constellations like GPS, Galileo, or GLONASS to calculate a location). Each receiver independently solves for its position and also reports an Integer Ambiguity Resolution (IAR) status. IAR refers to the process of figuring out exactly how many full satellite signal wavelengths fit between the satellite and the receiver, a step that, when resolved, produces a very precise position fix. A status of "fixed" means the receiver believes it has solved that count definitively.
The patent's key move is comparing the two receivers' outputs:
- Calculate the horizontal offset between the two reported positions.
- Compare that offset against the known antenna baseline vector.
- If both receivers report "fixed" confidence but the offset between their reported positions is larger than a threshold, something is wrong.
When that mismatch occurs, the system generates a wrong fix indication, alerting upstream software that at least one receiver's confident answer should not be trusted.
What this means for precision navigation in cars and drones
Precision GPS is increasingly important in situations where being even a few meters off is dangerous, including self-driving cars, delivery drones, and surveying equipment. The failure mode this patent addresses is particularly nasty because a wrong fix isn't a missing signal; it is a signal the system believes in fully. Current single-receiver systems have no reliable way to catch that kind of error on their own.
By building the cross-check into hardware that Qualcomm's long investment in positioning chipsets already points toward, this approach could raise a safety floor for any device running two antennas. For consumers, that could eventually mean a navigation system that knows when it doesn't know where you are, rather than driving you confidently into a wall.
That makes this Qualcomm's 34th filing we've tracked in our self-driving sensing watchlist since July, adding to fusing 3D sensors and cameras and one that pushed LiDAR range further.
The core tradeoff here is cost and complexity for reliability. Two GNSS receivers, two antennas, and the processor overhead to run both simultaneously add weight, power draw, and bill-of-materials expense to any device that adopts this approach. That is a real cost, and it probably keeps this design out of budget phones for a while.
But the thing the system gives up to achieve that reliability is minimal: the two receivers need to be physically close enough that their separation is known and stable, which is already true in most rigid-body platforms like car roof racks or drone frames. For those use cases, the tradeoff reads as clearly worth it.
What the patent does not address is what happens after the wrong fix flag is raised. Flagging bad data is only half the job; the harder question is what the device does next. That gap is either a limitation of this filing's scope or work left for a follow-on patent.
There are more where this came from
We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.
The drawings
13 drawing sheets from US 2026/0267013 A1 · click any drawing to enlarge
Want this weekly breakdown for a company we don't cover? Patentlyze Pro →