Improvement of GPS accuracy (30cm) with L5 frequency devices

For all high tech electronic equipment including GPS, PLB, chargers, phones, computers, software. Discussion of simple electrical devices such as torches, belongs in the main 'Equipment' forum.

Improvement of GPS accuracy (30cm) with L5 frequency devices

Postby keithy » Tue 13 Mar, 2018 5:35 pm

GPS Accuracy improvement
Late last year (2017), Broadcom - a manufacturer of GPS chipsets introduced the first dual frequency Global Navigation Satellite System (GNSS) chipset. This dual frequency receiving chip has the potential of improving consumer GPS accuracy from 5m to 30cm.

A Quick Background.
Consumer GPS and other GNSS rely on receiving a signal from orbiting satellites using a frequency known as the L1 Signal - 1575.42 MHz (10.23 MHz × 154).

Your GPS receiver uses this satellite signals to fix its distance from each satellite based on how long it takes the signal to go from satellite to receiver. It receives a minimum of three different satellite signals then computes your terrestrial position.

This is the frequency that GPS has operated on for over 2 decades. The addition of the L5 frequency band is part of the modernization of the GPS system.

This newer generation L5 frequency broadcasts consists of more complex signal at a different frequency and will be broadcast in conjunction to the legacy L1 signal. The L5 frequency has a higher transmission power, wider bandwidth and improved signal structure.

The L5 frequency will not only be used by the GPS system, but also by other GNSS in orbit - the European Galileo, and the Indian IRNSS and Japanese QZSS satellite systems.

OK. So now what?
Currently when a satellite's L1 signal is being received by a GPS receiver, the signal can reach the receiver directly, or indirectly by bouncing off buildings or canyons. The receiver calculates it's position based on the time that these signal took to reach the receiver from the satellite. As the direct and bounced signals arrive at different times and overlap, it results in lower accuracy as the receiver is looking for the peak of the signal to fix the signal arrival time. See the image below.

GPS L5 signal.jpg
GPS L5 signal.jpg (57.12 KiB) Viewed 4402 times

The newer dual frequency L1/L5 receiver will first lock on to the L1 signal, and then refine its positional computation using the L5. The L5 signal is so sharp that L5 signals don’t overlap with their reflections, so receivers can easily find the signal that arrives first. The receiver chip can effectively ignore any L5 signals after the first one it receives.

Why is this a thing now?
Up to now, there haven't been enough L5 capable satellites in orbit. 2015/2016 saw the addition of several L5 frequency capable satellites to earth's orbit. While the dual-frequency GPS has been used in commercial/industural applications such as in the oil and gas exploration industry, it hasn't been adopted in consumer based devices.

However, now there are enough orbiting satellites sending the both the L1 and L5 frequency (about 30 in orbit), with around 6-7 being visible at a position on a large portion of the globe, that it will be sufficient for mass market use.

What does this mean?
  • Greater accuracy - Using the L5 signal, it can mean that accuracy can be improved to around 30cm compared with the current 3-5 meters with the L1 Signal
  • Improved accuracy in cities, canyons and open areas - The L5 signal is less prone to reflections on buildings and other objects, so using both the L1 legacy signal and the L5 signal will greatly improve accuracy here.
  • Improved battery life - as a result of improved accuracy "searching" for satellites, the receiver should consume less battery power in achieving a GPS lock. Broadcom are indicating that their dual frequency GNSS chipset will use less than half the power of their previous GNSS receivers. "current consumption during GNSS tracking can be lower than 5 mA,” claims Broadcom.
When will we get it?
Hopefully some phones and devices will start sporting dual frequency GPS chips later in 2018. For the big phone brands, they are predominantly using integrated SoC (System on Chip) designs which incoporate the GNSS with the CPU (central processing unit) and GPU (Graphics Processing Unit). But hopefully it dual frequency GNSS will feature in newer models.

For dedicated GPS manufacturers like Garmin it might take a while to incorporate this into their receivers. They have been outpaced by smartphones for a while now. The consumer handheld GPS receiver market seems to be a fairly small market segment, and manufacturer's focus appears to be on their GPS watch range and their auto products. I still hold out hope that they incorporate these features into their devices sooner than later.
User avatar
keithy
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 625
Joined: Tue 28 Oct, 2014 5:31 pm
Region: Other Country
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby michael_p » Tue 13 Mar, 2018 9:36 pm

WOW! Can't wait for this to be available. Thanks for posting.

Michael.

Edit: some more info on the various new signals from gps.gov: https://www.gps.gov/systems/gps/moderni ... ilsignals/
Last edited by michael_p on Wed 14 Mar, 2018 7:55 am, edited 1 time in total.
One foot in front of the other.
User avatar
michael_p
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 987
Joined: Sun 15 Nov, 2009 6:58 pm
Location: Macarthur Region of Sydney.
Region: New South Wales
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby crollsurf » Tue 13 Mar, 2018 11:05 pm

Sorry but this technology is very dangerous. Useful for targeting someone in a crowd with a drone but offers little benefit to bushwalkers.
crollsurf
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 527
Joined: Tue 07 Mar, 2017 10:07 am
Location: Sydney
Region: New South Wales
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby highercountry » Wed 14 Mar, 2018 4:29 am

crollsurf wrote:Sorry but this technology is very dangerous. Useful for targeting someone in a crowd with a drone but offers little benefit to bushwalkers.


Paranoid or what?
Big Brother took over about three decades ago. Get with the program. :D
I can see many useful consumer applications.
highercountry
Athrotaxis cupressoides
Athrotaxis cupressoides
 
Posts: 196
Joined: Tue 19 Apr, 2011 8:52 am
Region: Victoria
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby Orion » Wed 14 Mar, 2018 6:11 am

Do you know how much the L5 will improve elevation accuracy?
Orion
Athrotaxis selaginoides
Athrotaxis selaginoides
 
Posts: 1713
Joined: Mon 02 Feb, 2009 12:33 pm
Region: Other Country

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby neilmny » Wed 14 Mar, 2018 7:28 am

crollsurf wrote:Sorry but this technology is very dangerous. Useful for targeting someone in a crowd with a drone but offers little benefit to bushwalkers.


The GPS locates itself based on receiving these signals.
Does the GPS actually talk (identify itself) to the satelites?
I'd like that accuracy in a PLB.
User avatar
neilmny
Lagarostrobos franklinii
Lagarostrobos franklinii
 
Posts: 2316
Joined: Fri 03 Aug, 2012 11:19 am
Region: Victoria
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby Mark F » Wed 14 Mar, 2018 10:30 am

Thanks for the heads up on the new technology keithy. I won't be buying any gps gadgets until the new tech starts appearing in them.
"Perfection is attained not when there is nothing more to add, but when there is nothing more to remove".
User avatar
Mark F
Athrotaxis selaginoides
Athrotaxis selaginoides
 
Posts: 1894
Joined: Mon 19 Sep, 2011 8:14 pm
Region: Australian Capital Territory
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby wander » Wed 14 Mar, 2018 11:00 am

Thanks for the heads up and the explanation. I often have my GPS accurate to a meter out in the bush (when you step from wiggle through scrub into a 2m x 3M tent site you took a GPS location from 3 years previous your know the system is accurate) and now I know that is because there is less reflected signals.
wander
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 651
Joined: Mon 26 Oct, 2009 11:19 am
Region: South Australia
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby Warin » Wed 14 Mar, 2018 11:16 am

neilmny wrote:Does the GPS actually talk (identify itself) to the satelites?


No. GPS receivers do not transmit.

There are a few control centers that do transmit to the satellites .. 5? around the world. But none of the normal GPS ones transmit to the satellites.

neilmny wrote:I'd like that accuracy in a PLB.


Why? If the rescuers are directed to within 20 meters of you that should be good enough.
Consider a PLB without the GPS function .. the search area IIRC would be 10 square kilometers...

This new stuff ... only a few of these satellites are up - so coverage will be poor.
The new satellites may not cover our area well, so may not have much impact on the GPS accuracy.
And a GPS that receives this extra band may consume more battery power, and be more expensive.
User avatar
Warin
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 531
Joined: Sat 11 Nov, 2017 8:02 am
Region: New South Wales

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby neilmny » Wed 14 Mar, 2018 2:20 pm

Warin wrote:............Why? If the rescuers are directed to within 20 meters of you that should be good enough.............


That (20 metres) would be in ideal conditions Warin.

As for the GPS does not transmit....that was my point operating a GPSR does not pinpoint your position.
Only a transmittier can do that. A smart phone locates you because it talks to the system even when you are not talking.
User avatar
neilmny
Lagarostrobos franklinii
Lagarostrobos franklinii
 
Posts: 2316
Joined: Fri 03 Aug, 2012 11:19 am
Region: Victoria
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby keithy » Wed 14 Mar, 2018 4:04 pm

Orion wrote:Do you know how much the L5 will improve elevation accuracy?

I would say that the L5 signal would improve vertical accuracy where that is impacted by multi-path errors (the reflected/bounced satellite signals). But the geometry issue still remains.

The basic gps spec for unaided signal observations (without ground or space based augmentation) and the global coverage being at least 4 satellites being available at a >5° elevlation at 99.99% of the time.

To get 3D gps fix you need a minimum of 4 observable satellites - more for greater vertical accuracy. GPS accuracy is still greater horizontally than vertically. This is due to the curvature of the earth. To better explain I drew up this pic:

GPSSatellites.jpg
basic GPS diagram by keithy


The dual L1 & L5 signal can also improve the signal by being able to estimate the ionospheric signal delays directly (without the augmentation by WAAS or SBAS). The stronger L5 signal should also reduce noise and multi-path errors to give a more accurate vertical accuracy.

wander wrote:now I know that is because there is less reflected signals.

There are also issues with the tropospheric and ionospheric conditions that can delay the signal transmission. This might also decrease your accuracy. That is why augmentation systems like WAAS exist. Unfortunately, we don't have any ground based GPS augmentation stations in Australia.

Warin wrote:Why? If the rescuers are directed to within 20 meters of you that should be good enough.

That 20 m is the best case using existing GPS technology. Obviously challenging locations (canyons/mountainous regions/dense tree canopies), this can negatively impact on the GPS accuracy. The L5 signal has stronger transmission power, and more robust redundancies, so should improve accuracy if used in PLBs as well.

Warin wrote:This new stuff ... only a few of these satellites are up - so coverage will be poor.

The first L5 capable satellite went up in 2009, and now there are 30 L5 capable satellites, including one that orbits Japan and Australia only, and so for most of the world, there will be 6 or 7 satellites visible. Remember that the GPS constellation is 24 operational satellites. This does not include the non-US GNSS like Glonass etc. If you have a GPS receiver or a GPS capable phone and a satellite app, check out your satellite page, and if you see any of these satellites 01 03 06 08 09 10 24 25 26 27 30 32 - these are capable of L5 transmission. The Indian IRNSS and the Japanese QZSS in orbit now also transmits L5. The Japanese satellites that can are 193 194 195 199.

Warin wrote:And a GPS that receives this extra band may consume more battery power, and be more expensive.

Broadcom released their spec of their dual frequency chipset in September 2017 and it shows it consumes less than half the power of the previous generation Broadcom GNSS chipset, as mentioned above "current consumption during GNSS tracking can be lower than 5 mA".

When asked about the price to OEM (original equipment manufacturers), Broadcom stated the price of the chipset would be the similar to the existing L1 GNSS chipset.

Of course, I wouldn't put it past a GPS manufacturer jack up the price of new models incorporating these dual frequency receivers with marketing spin.
User avatar
keithy
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 625
Joined: Tue 28 Oct, 2014 5:31 pm
Region: Other Country
Gender: Male

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby Warin » Fri 16 Mar, 2018 5:00 pm

keithy wrote:
Warin wrote:And a GPS that receives this extra band may consume more battery power, and be more expensive.

Broadcom released their spec of their dual frequency chipset in September 2017 and it shows it consumes less than half the power of the previous generation Broadcom GNSS chipset, as mentioned above "current consumption during GNSS tracking can be lower than 5 mA".

When asked about the price to OEM (original equipment manufacturers), Broadcom stated the price of the chipset would be the similar to the existing L1 GNSS chipset.

Of course, I wouldn't put it past a GPS manufacturer jack up the price of new models incorporating these dual frequency receivers with marketing spin.


That is all good stuff.. was not aware they had progressed so far with the L5. I'll keep my old GPS going just the same ... the longer I wait the better the eventual replacement. Reasonably happy with my old one - ~10 meters is enough for me to find stufff.
User avatar
Warin
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 531
Joined: Sat 11 Nov, 2017 8:02 am
Region: New South Wales

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby Orion » Sat 17 Mar, 2018 1:58 am

keithy wrote:
Orion wrote:Do you know how much the L5 will improve elevation accuracy?

I would say that the L5 signal would improve vertical accuracy where that is impacted by multi-path errors (the reflected/bounced satellite signals). But the geometry issue still remains...

...The dual L1 & L5 signal can also improve the signal by being able to estimate the ionospheric signal delays directly (without the augmentation by WAAS or SBAS). The stronger L5 signal should also reduce noise and multi-path errors to give a more accurate vertical accuracy.


Yes, thank you. I understand the basic geometry. But my question was "how much" it will improve the vertical accuracy. Will it scale similarly to the improvement in horizontal accuracy?
Orion
Athrotaxis selaginoides
Athrotaxis selaginoides
 
Posts: 1713
Joined: Mon 02 Feb, 2009 12:33 pm
Region: Other Country

Re: Improvement of GPS accuracy (30cm) with L5 frequency dev

Postby keithy » Tue 20 Mar, 2018 10:27 pm

Orion wrote: But my question was "how much" it will improve the vertical accuracy. Will it scale similarly to the improvement in horizontal accuracy?


I don't know the answer. Broadcom do not make their GNSS chipset accuracy data available, unlike the manufacturers who supply commercial/survey grade GNSS equipment. With current dual and triple frequency (L1/L2/L5) GNSS receivers used for survey work, the vertical accuracy in these receivers (without WAAS/SBAS or DGPS augmentation) is around 1.5 - 2.0 times worse than their horizontal accuracy. Eg. if the Horizontal Accuracy 1m / Vertical Accuracy 2m.

While I don't expect that the consumer use chip will be as accurate as commercial/survey grade equipment, I expect that the vertical accuracy improvement to be in some similar scale using the L5 frequency.

Of course, you will still have errors due to the difference between GPS elevations being based on an ellipsoid, while maps elevations based on a vertical datum tied to the geoid.
User avatar
keithy
Phyllocladus aspleniifolius
Phyllocladus aspleniifolius
 
Posts: 625
Joined: Tue 28 Oct, 2014 5:31 pm
Region: Other Country
Gender: Male


Return to Techno-Babble

Who is online

Users browsing this forum: No registered users and 4 guests