Why GPS Fails In Cities. And What Researchers Think Could Fix It

ScienceDaily reports: Our everyday GPS struggles in "urban canyons," where skyscrapers bounce satellite signals, confusing even advanced navigation systems. Norwegian University of Science and Technology (NTNU) scientists created SmartNav, combining satellite corrections, wave analysis, and Google's 3D building data for remarkable precision. Their method achieved accuracy within 10 centimeters during testing [90% of the time]. The breakthrough could make reliable urban navigation accessible and affordable worldwide...

"Cities are brutal for satellite navigation," explained Ardeshir Mohamadi. Mohamadi, a doctoral fellow at NTNU, is researching how to make affordable GPS receivers (like those found in smartphones and fitness watches) much more precise without depending on expensive external correction services.

Re: Solution in search of a problem

[idontusenumbers]

You're probably utilizing a secondary location service like wifi or tower based location, which isn't as accurate as this claims to be.

Re: Solution in search of a problem

[kellin]

You've clearly never walked around Manhattan or any city with massive skyscrapers. This *is* a problem. Hell, there are spots in LA where there aren't skyscrapers and my GPS mapping fails.

Re:

[_merlin]

I've used the GPS on my DSLR in Hong Kong and never had an issue getting a fix anywhere outdoors, even with skyscrapers on both sides of the street.

Re:

[piojo]

No, it's not a problem getting a fix. It's a problem where you appear on the opposite side of the street due to signal bounce.

Re:

[Ryanrule]

Mine car systems and phones have worked just fine in downtown chicago.

Re:

[s0nicfreak]

When you're driving, you don't notice because you're moving fast (which means it gets more data) and hopefully not looking at the map (so you don't necessarily notice if it's inaccurate for a moment). You're also bigger, likely to be on one side of the street depending on which direction you're traveling (so if it just estimates, it's usually right), etc.

I've certainly noticed a problem in downtown Chicago while walking and biking. (In other places with skyscrapers too, but I'm in Chicago a lot so I noti

Re:

[strikethree]

Manhattan and Washington DC both have active jammers. It is not your device or the satellites that are at fault, it is the government denying you a fix nearby to something they think could be a target for precision weapons.

Hell, there are spots in LA where there aren't skyscrapers and my GPS mapping fails.

Then you are near something sensitive to "National Security". GPS is not as flaky as you appear to think it is.

(ROFLMAO, CAPTCHA is 'censored'. How the hell are CAPTCHAs so frequently insightful/accurate?)

Re:

[Tony Isaac]

Apparently you haven't done much driving among big city skyscrapers. GPS is very intermittent in large hubs of cities like NYC, Chicago, LA, Houston, and Dallas. Sometimes, when coming out of a parking garage, it can take minutes before the GPS picks up your location.

Re:

[tlhIngan]

It's not hard to see why - a GPS satellite isn't a really powerful thing - think of it as just a lightbulb transmitting signals and you're basically seeing how strong the signal is. It's basically in the noise by the time it reaches earth.

The other problem is the signal geometric dictates how accurate it is - a city with tall buildings means you only have a very narrow view of the sky, and that narrow view leads to large GPS errors. It's why everyone uses wifi and tower trilateration to get a better positio

Re:

[registrations_suck]

Then it must not happen to anyone else either and the article is a fucking lie!

Re:

[Computershack]

Ddisable Wifi and Bluetooth scanning in Location Services then tell me how well you get on.

Re:

[volcan0]

I would have had the same reaction as you normally. I live in a big city ( on of the biggest of the country ) and never experienced this issue. However, we went to New York recently and GPS was basically unusable downtown (midtown).

Re:

[2TecTom]

Me either and I drove professionally, when I do see people having problems with GPS, they are often people not paying attention, misusing the technology or not understanding its limits. On the other hand at lot of development these days is crap, management is so often incompetent too.

It's about time

[Ronin Developer]

We were in NYC the other week. Even walking through Central Park and using Apple Maps was a trial. It frequently misreported our position on the trails.

In the city proper, it was far worse.

Normally, Maps, Google Maps, and Waze are spot on...just not in NYC.

Re:

[stealth_finger]

We were in NYC the other week. Even walking through Central Park and using Apple Maps was a trial.

You need gps for that?

Re:

[Anonymous Coward]

Person A: I can't get good GPS location data while in city streets with tall buildings around me. My locations is sometimes two or three streets off!

Person B: There is a new method of analyzing the phases of the signals in those multipath environments to eliminate the reflections and duplications. It can fix your position down to 10cm.

Person A: But the Ruskies can use that to pinpoint missiles at me!!

Person B: Yeah, the Ruskies need to worry about missile guidance between buildings in city environment

Re:

[russotto]

When the Ruskies need to hit a 2m MTA exhaust port located among skyscrapers to blow up the city, this technology is really going to help them a lot.

Re:

[anoncoward69]

The Ruskies don't need this if they want to nuke NYC. They'll just air burst a nuke half a mile above NYC where it still has line of sight to the GPS sats that it needs and can do maximum damage.

NOT GPS its FUSION - real solution QZZ

[johnjones]

you can do offline processing and fusion of signals but really

Quasi-Zenith Satellite System (QZSS), nicknamed "Michibiki" is the best solution hey if you had enough LEO broadcasts you would be good but starlink does

Combining different GNSS systems is also an option

[williamyf]

Chances are that, when the GPS satellite is bouncing in the urban canyon, the Galileo one is in the zenit. And if you live in many countries where geopolitics are an aftertought, you can use GPS, Galileo, Glonass, and beidu, all at the same time. what's not to like?

Polish scientists have been doing this since 2018, probably there are even older efforts:

https://www.mdpi.com/2072-4292... [mdpi.com]

You combine that with cell tower location data, which is cheap, and Robert's your uncle.

Re:Combining different GNSS systems is also an opt

[monkeyxpress]

Polish scientists have been doing this since 2018

Um...commercial receivers (e.g. anything from ublox) have been doing this since at least 2010 (when I started using them).

The standard positioning signal on all of those GNSS systems is not encrypted. Your front end is already picking it up, so you might as well process it, whether you like a country or not, and that's exactly what they do.

The issue with urban canyons is not that there aren't enough satellites (sure that helps) but that you don't have a wide enough view of the sky for accurate triangulation. If you can only see a small sliver of sky directly above you, the trigonometry means your accuracy is going to be limited. Add in problems of reflections of those buildings and you've messed up all the path lengths and now your result are going to be all over the place.

This looks like a very impressive solution.

Re:

[azouhr]

Add acceleration sensors to the equation, they are really precise today. Those things can navigate for a certain range without any satellite signal. The other thing that cars sometimes do, is measuring the distance and they also know about the direction the cars go from the steering angle.

Re:

[timeOday]

Isn't an INS of that quality pretty expensive? I know, every smartphone has an accelerometer, but I don't think they're navigation-grade.

Re:

[coofercat]

Indeed, getting line of sight when there's a 50 storey obstruction next to you is going to be tricky. So is signal bounce. GPS receivers don't really have the concept of "signal strength". That is, the GPS signal is below the noise floor anyway, so you can't really measure the strength of it. Instead, you have to spread-spectrum hunt for the edges of the digital signal, and when you've got a few edges in a row which look like they match the pattern you're expecting, then you declare that signal "locked on"

Re:

[caseih]

My Pixel 4A is ancient now and it has always fixed with all the constellations. I think at any one time it's seeing 30 odd satellites, all contributing to the fix.

Re:

[thegarbz]

Polish scientists have been doing this since 2018

Dude common off the shelf phones have done this since 2012.

Next, compass please

[Fons_de_spons]

Great! Next, a built in compass that actually knows where north is... No matter how many times I calibrate mine, it is always somewhere else. On all smartphones I ever owned...

Re:

[tudza]

I used to carry a compass and navigated with a static Google map. That was San Fransisco. Using just Google maps and GPS I did find I'd have to walk a certain distance along a block before being reasonably sure I was headed in the right direction. That was NYC.

Re:

[Mspangler]

Given that the magnetic field strength is falling and magnetic north is moving much faster than usual as the Earth's magnetic field gets ready to flip that might not be possible just now (where just now refers to the geological version).

The current North Star is still good for a few thousand years though but precession will eventually move that to a different star.

Re:

[thegarbz]

The electronic compass in your phone works perfectly fine. It's as reliable as any magnetised needle. The problem is it is built into a device which interferes with itself by necessity (you can't shield the compass without breaking it's functionality, you can't stop the EM interference without breaking the other functionality of the device).

You're asking scientists to develop a compass that doesn't rely on the very thing compasses use to work. If you need a reliable one, buy literally any compass (digital

Re:

[anoncoward69]

Exactly, go get an old school magnetic compass and set it on top of your phone and tell us what happens? The built in compass chips have to attempt to cancel out the EMF and magnetic fields from the phone it's hosted in. That's why you usually need to move the phone in a figure 8 pattern every now and then to calibrate that cancellation to make the compass more accurate.

Available for years

[gnasher719]

At least Apple claimed years ago that iPhones used GPS when available and used a database of WiFi network locations otherwise. Because in the locations where you donâ(TM)t get GPS you typically get tons of WiFi signals.

Really?

[nospam007]

"Their method achieved accuracy within 10 centimeters during testing [90% of the time]. "

Yes, the 90% of the country where there are no cities.:-)