Super-Accurate GPS Chips Coming to Smartphones in 2018
spectrum.ieee.org
spectrum.ieee.org
Didn't the US scramble the civilian signals specifically so it wouldn't be too accurate? And they're just okay with civilians having access to ultra accurate GPS receivers?
Depending on the source of signals used, the military can also just change the precision for civilian bands. But in times of peace / low threat levels, there's not a particular need to do so.
http://lea.hamradio.si/~s53mv/navsats/theory.html
No FPGAs, RPis, or SDRs --- it's mostly done in discrete logic with a 68000 doing the back-end processing. As a bonus, he also implemented GLONASS. Furthermore, the software takes up less than 32KB of ROM.
But practically speaking, almost all GPSes are made outside of the US and then imported...
See page 5 of DTIC document: Relativistic Effects in the Global Positioning System
http://www.dtic.mil/dtic/tr/fulltext/u2/a158720.pdf
Without correction the drift due to combined coarse relativistic effects would be on the order of 100,000 nanoseconds per DAY.
In fact the first NTS-2 satellite was launched with the clock unadjusted for GR, due to doubters, but after three weeks the syntheziser had to be activated to compensate. The calculated uncorrected drift on that clock alone was 38,000 nanoseconds per day.
As the document notes, to further refine Navstar to greater levels of precision would require tackling yet another set of tougher relativistic effects, plus more mundane atmospheric influences.
GPS satellites are all tuned this way, and it has the desired effect.
However, for people implementing GPS receivers, the signal arrives already corrected for time dilation. So the people making that part of the system don't have to know about general relativity at all :)
Also gives you the point - if one guy can build a receiver on his own... I mean, I don't want to make it easy for rogue nations and terrorist groups to get ICBM-mounted nuclear weapons, but with the difficulty of building and testing the missiles and bombs, it seems a little silly to expect restricting GPS receivers to accomplish anything.
By default all GPS units have some built-in limits (so-called COCOM
limits) that prevents them from providing any data if the velocity
exceeds 515 m/s at altitudes above 18 km. These artificial limits are
built into GPS receivers to prevent bad guys from using them in
missiles and other nasty stuff.
I was confused. So here's a translation: Bad guys = US enemies or non-allies. Not bad guys in any moral sense.
Or at least that's how I always understood it.
Makes them worthless for high altitude near space balloons :-(
> In May 2000, at the direction of President Bill Clinton, the U.S. government ended its use of Selective Availability in order to make GPS more responsive to civil and commercial users worldwide.
> The United States has no intent to ever use Selective Availability again.
The military has better precision, only because they buy better hardware. Similar hardware is used in commercial drones to improve GPS accuracy.
> Using two GPS frequencies improves accuracy by correcting signal distortions caused by Earth's atmosphere. Dual-frequency GPS equipment is commercially available for civilian use, but its cost and size has limited it to professional applications.
That's not true. Military receivers can use the Y Code which is encrypted for military use only. There are some really fancy civilian receivers out there that can leverage Y code to a limited extent without knowing the encryption key but full accuracy is only possible on a military receiver that can decrypt and use the Y code.
Knowing the lane of the highway you are may be useful as well.
Obviously the self-driving cars of the future will also need other sensors as GPS doesn't work under tunnels and bridges.
[1] https://agriculture.trimble.com/precision-ag/products/steeri...
Most if not all airliners equipped with autoland systems use ILS rather than GPS for position as well.
Yesterday I read an idea from a guy who wanted to create a mariokart-style "ghost" via AR to pace you while you are on a run. That's one example.
https://en.wikipedia.org/wiki/GPS_signals#Military_.28M-code...
Though I'm honestly surprised I've "known" of GPS for this long and never realized there were a variety of codes. Legacy--some still operating, modern, military, and more. I'm going to enjoy binge reading this wiki/math stuff tomorrow.
Also, I've been wondering, how effective the old scheme with a secret military and a public code was - certainly an old GPS reciever shows up sometimes at a military surplus store. I wouldn't be surprized if enemy nations bought one of those, and extracted the secret parameters from the now quite old chips using modern technology.
E.G.
P code 1001011010010110101010100010101
W code 00001111000000001111111100001111
And since it's modulated with the W code this would leak information. Not really sure how off base that part is, that's just a rough explanation I was told many years back.
Semi-codeless receivers can use it anyways. The reason is that the W code has a much lower data-rate (500KHz I think).
Take the incoming signal, correlate with the aligned P code. The resulting signal will be a MHZ wide signal, bandpass to it, and square the result. The squaring wipes off the residual W code. Because you bandpass filtered before the squaring you don't get so much squaring loss. Lock the resulting carrier, and use RTK signals and long observations to resolve integer ambiguities.
For a dual frequency rx you can exploit that P(Y) is the same on both frequencies and correlate them against each other to find the ionospheric delays.
Would you have links to any examples?
Some will even switch of the radio once one of these limits are reached and not enable it until it's been reset so make them a pain to use for even if you just wanted the location data after your payload comes back down within the limits.
EDIT: But yeah the so-called COCOM limits still apply to civilian equipment (esp to a company such as Broadcom). Unsure on the regulations for modifying the firmware on a device that doesn't put these lockouts at the silicon layer. For example: https://github.com/swift-nav/libswiftnav/blob/master/src/pvt...
The original limits appear to be for units available for export, but I would guess manufacturers didn't want to deal with the headache of producing different models for import/export and just slapped the restriction down on everything. The poster does claim the original restriction was indeed an OR[1].
And current restrictions may only be 600 m/s? [2]pg 58
[0]https://space.stackexchange.com/questions/14687/current-situ...
[1]https://www.ngs.noaa.gov/corbin/class_description/5700-5800V...
"Immediate access to satellite measurements and navigation results is disabled when the receiver’s velocity is computed to be greater than 1000 knots, or its altitude is computed to be above 18,000 meters. The receiver continuously resets until the COCOM situation is cleared."
[2]http://mtcr.info/wordpress/wp-content/uploads/2017/07/MTCR-T...
That kind of limit is implemented in firmware (all the math is done in software), and they aren't generally protected beyond being an obscure architecture. When you've decided whatever vliw dsp architecture they use, finding a simple 'if' condition near the output logic is quite easy.
https://en.wikipedia.org/wiki/Error_analysis_for_the_Global_...
I guess I'd reverse the question. Why wouldn't you want the greatest level of accuracy you can get? Maybe you want to track your property lines? Maybe you want to accurately mark where your access point is for your septic tank? Maybe you just want to hide Easter eggs or buried treasure?
I'm not sure why anyone would want a tool that is less accurate than it can be, assuming it is also reasonably affordable.
When moving slowly (walking).
When precise coordinates are needed (self driving vehicles).
When precise altitude is needed (airplanes and drones).
When tracking smallish things on a person (head tracking for augmented reality).
When performing certain scientific measurements (continental drift).
Finally, so that when I'm sitting at a stop light my navigation doesn't freak out and suddenly think I'm moving in the opposite direction.
I'm sure there are tons more.
If I understand correctly, the best implementations treat the various satellites as one combined system, rather than trying to average the solutions.
I do see the point of the intention, but as far as I understand it, it never really worked for preventing that anyway.
Maybe try telling it that you're biking instead, my guess is they do more smoothing the higher speed they expect.
Chips with L5 should do a lot to clear that up though.
A GPS unit is useful for assisting dynamically changing routes (e.g. driving) or to track unknown paths (e.g. plotting your hike in snow/darkness). May I ask what you'd need a fine-accuracy GPS unit for during a run?
I mostly find the GPS invaluable in hiking. It can be a great thing to have on the top of a mountain in the winter with low visibility. If you get up to a summit and the weather turns, you can have the watch send you back down the way you came. Nice and safe.
On a longer run they are probably accurate to within 1% which makes practically no difference but you often find on a track that the inaccuracy multiplies on each lap so if you are doing say 1 mile repeats (4 laps) then by the end of the mile you can be 20 metres out.
It's not a biggie because visually you can see if your watch says 0.23, 0.46 etc as you go past the 402m mark you know that you need to up your pace because it's short but having a more precise distance and therefore pace would save a few brain cells when you're trying to concentrate on hitting your paces.
With the existing L1 signal the multipath signals overlap each other and its hard to find out which the original signal was.
The newer L5 signal is much shorter, so multipath signals will show up as individual peaks, and the receiver can simply pick the first signal, cause that will be the direct one.
What makes anyone confident in the accuracy of wearable tracking devices? I haven't seen it verified and I wouldn't assume it: Accuracy is expensive, generally speaking, and few consumers will pay for it or even question it - how often have you heard someone mention it?
Or, for example, have you wondered how accurate your simple bathroom scale is? I looked into it a little, wanting to collect accurate health data: IIRC +/- 0.1 kg (~0.2 lbs) is available in <$100 scales, but for real precision you need the $500 scales at your doctor's office. And how consistent is it? I tested mine, a decent one with good reviews, and getting it to report the same weight in immediately consecutive measurements was a challenge; I had to stand on it in just the right way.
Here are a couple articles that found the accuracy of activity trackers wanting. I've seen other articles in places like ZDnet that found the same problems, though based on less rigorous research.
Accuracy of Smartphone Applications and Wearable Devices for Tracking Physical Activity Data in JAMA <https://jamanetwork.com/journals/jama/fullarticle/2108876>
Accuracy Of Fitbit Data Seriously Questioned By New Study <https://www.sporttechie.com/accuracy-of-fitbit-data-seriousl...
(Sorry if this disrupts your confidence in your data!)
I tested our old Nintendo wii (or whatever it's called) balance board and so far it gives the same value every time I step on it. The value was also only 0.2 kg off from the doctors expensive scale.
Yeah, I noticed. Does anyone know how this data is normalized by professional researchers?
“What the plaintiffs’ attorneys call a “study” is biased, baseless, and nothing more than an attempt to extract a payout from Fitbit. It lacks scientific rigor and is the product of flawed methodology. It was paid for by plaintiffs’ lawyers who are suing Fitbit, and was conducted with a consumer-grade electrocardiogram – not a true clinical device, as implied by the plaintiffs’ lawyers. Furthermore, there is no evidence the device used in the purported “study” was tested for accuracy.”
Fitbit’s research team rigorously researched and developed the technology for three years prior to introducing it to market and continues to conduct extensive internal studies to test the features of our products. Fitbit Charge HR is the #1 selling fitness tracker on the market, and is embraced by millions of consumers around the globe.
Consumer Reports independently tested the heart rate accuracy of the Charge HR and Surge after the initial lawsuit was filed in January and gave both products an “excellent” rating. We stand behind our heart-rate monitoring technology and all our products, and continue to believe the plaintiffs’ allegations do not have any merit. We are vigorously defending against these claims, and will resist any attempts by the plaintiffs’ lawyers to leverage a settlement with misleading tactics and false claims of scientific evidence.
There are some foot pods that will outperform GPS.
All current running watches and phone software uses single point calibrations. Now if they curved fit 2 or 3 points....
On my older iPhone, the mile counts were reasonably accurate. On the 7, there's a spurious ~2 miles per hour added at all times.
My guess is that the 7's chip updates position much more frequently, so that small inaccuracies (a couple dozen feet west, then a couple dozen feet east on the next update) make it look like I'm sprinting all over the place in random directions at high speed. No idea why Fitbit doesn't smooth it out a little...
Did a six hour, 4.6 mile hike in the Adirondacks and it tracked me at 18 miles (edit: 13, sorry).
If the terrain forced you to take a lot of shorter steps then that explains it. In 6 hours of normal steps you'd expect to walk about 18 miles.
Standing still for a break still makes my mileage go up, and I can see the blue dot on the map in the Fitbit app jitter around a little - a few feet here, a few feet there.
Here's an example of the generated map: https://imgur.com/a/yoqei
(That's a ~5 mile hike in actual mileage - and you can see the issue very clearly between "miles" 4 and 5 on the map, where we likely took a long break, and in the longer distance between the downhill miles as we were moving faster)
(100 meters isn't that far; 1 cm at that range is already 1/10000 of the distance; Commercial equipment can apparently do millimeters across kilometers: http://www.micro-epsilon.com/displacement-position-sensors/l... )
...and I'm surprised by the silly choice of words in English: it's really confusing.
And yeah, as a test engineer, metrology can be confusing, but that's because it's necessarily complex, not because of silly choices.
The short answer is no, largely due to localised ionospheric conditions. The longer answer is yes, GNSS systems can be enhanced in accuracy using RTK [0][1]. This allows for precision on the order of 1cm, but requires a (rather expensive) fixed position base station to provide corrections. Surveyors, precision ag, and other large machine control (think dozers, graders, etc) have used this technique for precision positioning for the past decade or so.
I do recall reading that upcoming GPS satellites add a second civilian frequency, which I believe will enable these sort of ionospheric corrections to be done more easily and inexpensively.
The antenna systems on RTK rovers (the devices receiving corrections, which also receive the GPS satellite signals) aren't huge, but I'm not sure if they could be reduced in size enough to fit in a smartphone.
https://www.swiftnav.com/sites/default/files/whitepapers/loc...
The technology in good receivers like the Piksi is slowly drifting into consumer-level hardware. You need to correct for such errors as:
1. Jitter in the code phase vs. phase-locked-loop. As the article describes, most existing codes are long and slow, with bandwidths on the order of 300 m (specifically, 1.023 MHz). The receiver can measure the exact time at which this code changes to get your existing resolution, but the L5/E5 bands are at 20 MHz.
2. Integer offsets when interpreting the carrier frequency. The above slow codes are transmitted on a carrier at 1-2 GHz, giving much higher resolution, but there's no information in the signal as to which particular node of the carrier frequency you're observing. Comparing to a known location, and improving your guess over time, can let you make use of this information.
3. Unknown, slowly drifting ionospheric delay. Typically fixed in industrial or agricultural applications by using a base station and radio link to tell your remote link that the base station (which is bolted to a big chunk of concrete) is now reporting that it's 20 cm from where it was an hour ago, and the remote unit should probably just adjust any measurements by that much.
I only have enough knowledge of the system to be dangerous, but I've wondered whether it would be possible to correct for #3 at a consumer level with a phone app. If you had thousands of phones cooperating in a city, at any given time many would be stationary, even charging or on wifi, and you could theoretically trade off roles as reference base stations and remote receivers. I think it would require a lot more low-level access to the GPS chip than generic Apple/Android phones give you, but it's an idea - feel free to take it and run if you like it.
To answer the basic question - common NMEA protocol returns GPS data with latitude formatted as DDMM.MMMMM (Degrees, minutes, and decimal minutes) and longitude in DDDMM.MMMMM format. Four digits of precision get you a precision of about 1.6 meters, depending on where you are on the Earth, but that doesn't mean you have that level of accuracy.
Indeed, carrier phase and multi-frequency+constellation measurements are drifting into consumer hardware, but I can say that it's probably going to be a Good Long While before we full realize its benefits. The use of cell phones (and cheap cell phone antennas) is an active area of research, particularly for organizations aiming to make high-accuracy GNSS positioning widely accessible outside of specialty markets (i.e., agriculture and surveying). It's an interesting area to be in for sure:
- http://gpsworld.com/innovation-precise-positioning-using-raw-gps-measurements-from-android-smartphones/
-
https://radionavlab.ae.utexas.edu/images/stories/files/papers/inPhoneCdgnssIonPlans2016.pdf
(Shill: We're hiring for firmware engineers right now, focused on a variety of different areas: embedded Linux, DSP basebands (C), and navigation algorithms (C++) design and implementation. See https://jobs.lever.co/swift-nav for more details or email jobs@swiftnav.com.)
Iono delay is typically corrected by using a dual-frequency receiver so that it can be calculated directly via difference in arrival times of the time pulse on each frequency. L2C can help here.
It is possible that a setup with high-precision absolute measurements can produce high-accuracy relative measurements, but the former does not imply the latter (imagine, for instance, any source of error in absolute measurements that's locally non-linear).
I wouldn't expect interesting oil/gas pockets to have <5m lateral diameter, but then again I know very little about that industry.
Also drilling is done to extremely high precision these days.
Edit: Bare in mind the industry uses differtial GPS for some tasks - this can give mm/cm precision.
There's a ton of money in oil/gas, and any technology that can give a slight edge is often worth it. For example, oil and gas companies were some of the first to use Iridium (worldwide satellite internet at ridiculous prices), and I'd bet that they are the biggest users of the Iridium network to this day. This goes for inventions in tons of fields.
For example, new breakthrough in diving: great, you can use that to work on underwater parts of drilling rigs. New breakthrough in robotics? Great, you can replace the expensive and dangerous dive crews with underwater ROVs. New breakthrough in geology? Cool, you can use it to find oil.
They're the lesser of two evils. Still an evil, though.
Is that even true? They pulled the plug on mobile ad sales a year or two ago (http://appleinsider.com/articles/16/01/18/after-abandoning-i...), where else would they make advertising revenue?
And in any case, a company who makes 99.9% of their revenue off of non-ad sales is definitely would never be categorized as an advertising company. My grocery store is paid to put crest at eye-level, that doesn't make them an advertising company...
IDFA is still going strong on Apple phones and is likely to continue into iOS 11.
I don't know if they're ads or not but there's also location based app suggestions too.
disclaimer: i'm a googler so glass house throwing stones etc, and of course the scale is way different
http://searchengineland.com/apple-expanding-successful-searc...
The difference between 5m accuracy and 30cm accuracy isn't terribly important if you're just trying to tell what building someone is in, and it won't improve in-building accuracy. I can't think of anything sketchy that it enables that someone couldn't do already.
Refining GPS accuracy only modestly improves the ability to track what stores you are near. Wifi-assisted GPS tracking that exists today probably tells advertisers everything they need to know already.
http://www.nytimes.com/2013/07/15/business/attention-shopper...
RTK is commonly used in survey drones, and kits are available for about $1K.
With the new chips, Broadcom, u-blox and others are bringing to market now you are no longer dependend on a base station to get cm-level accuracy. Instead they exploit the different properties of the L1/L5 frequencies to infer stuff about e.g. the atmosphere, as the different frequencies are altered in different ways while travelling to earth[2]. The system can thus reduce its margins of errors[1] in the position calculations. So they are not "interacting" with RTK in any way. The chips are also much cheaper than your quoted $1K price tag.
Disclosure: I work for u-blox, but not an expert in GNSS calculations
[1] https://en.wikipedia.org/wiki/Error_analysis_for_the_Global_... [2] https://www.e-education.psu.edu/geog862/node/1715
Whereas the dual frequency approach can be used in the same way as "regular" GNSS, using all the low-power tricks to sleep as much as possible (with some accuracy vs. power trade-offs). Of course there will be a power consumption penalty vs. a single frequency receiver: the two RF chains, and the extra base-band processing. The later can be mitigated by better nodes (the article mention the chip being 28nm, so low dynamic power). The two RF chains impact of course can't be avoided. But for some application it may be worth it.
[0] http://leica-geosystems.com/en-us
Even highly trained and skilled cane users would wander into the roadway usually within 10-15 minutes of using the system, purely because of GPS drift. And forget about ever using it in urban centers. In London, Boston and NYC it was rare to be more accurate than 50-100m. It's scary how bad GPS accuracy really is.
I frequently run my ten year old TomTom, alongside my wifes iphone or my android on road trips. The tomtom is pretty much always dead on with respect to lanes/turning/etc, while the iphone/android regularly feeds us crap. That said, I would kill for google maps (pc browser version) style route planning on the tomtom where you can click various roads and get time estimates even if goggle doesn't think its a good route.
Using the other constellations helps you see more sats, but it doesn't really help deal with the reflection issue in urban environments. The "big deal" about this chip is that is uses the new L5 signal (and it's equivalent in other constellations). Previously that has only been available in very expensive hardware.
The BCM47755 can simultaneously receive the following signals:
GPS L1 C/A
GLONASS L1
BeiDou (BDS) B1
QZSS L1
Galileo (GAL) E1
GPS L5
Galileo E5a
QZSS L5
> However, L5 signals are so brief that the reflections are unlikely to overlap with the direct signal. The receiver chip can simply ignore any signal after the first one it receives, which is the direct path.
Can someone explain this? Surely the first signal received will always be the direct signal, how could you receive signals from reflections first?
So yes, the direct signal always arrives first, but gets messed up during receiving by the reflected signal. The L5 signal is so short, that the likelihood of reflections overlapping is reduced. Think: so short that the signal starts and finishes during the time it takes a radio wave to propagate 1 meter through air or so.
Photogrammetry is what humans do, and it doesn't cost billions of dollars.
https://en.wikipedia.org/wiki/Global_Positioning_System#Deve...
The reason why I ask: If we can get highly accurate and signed timing data it will have a huge impact on distributed systems. If the signature is accessible from the API it could be included in a DB transaction that every node could verify.
It's awesome that it is half the power of current chips too. Because I think smart phone users care more about battery life than accuracy.
Edit: Changed my messages because I reread the article and caught something I missed the first time. I missed the power usage statement the first time.