IPhone 4S has GLONASS in addition to GPS | Hacker News Reader

IPhone 4S has GLONASS in addition to GPS

apple.com

129 pointsdchest89 comments

Precisions:

  * GPS: 2.00—8.76 m
  * GLONASS: 4.46—7.38 m
  * GPS+GLONASS: 2.37—4.65 m [citation needed]

(source: http://en.wikipedia.org/wiki/GLONASS).

From which:

    To improve development of the user segment, on August 11, 2010, Sergei Ivanov
    announced a plan to introduce a 25% import duty on all GPS-capable devices,
    including mobile phones, unless they are compatible with GLONASS.

I think I can see why the GPS chipsets are starting to adopt this...

That's a 25% import duty on devices imported into the Russian federal territory.

Yeah, sorry - that's not totally obvious from the bit I snipped. Apple (and presumably the other phone manufacturers) do sell into Russia, and a 25% import duty would probably have a significant impact on their sales.

I was surprised when I saw GLONASS support appear in recent Garmin devices. http://garmin.blogs.com/my_weblog/2011/09/garmin-etrex-and-g... Now it starts to really make sense.

That's hardly a motivation for Apple. They still don't have an official dealer in Russia and afaik don't have any plans. Yet, most of government officials widely use iPads (including the president) and every hip kid around the block has an iPhone. This suggests that Sergei Ivanov is merely looking for more ways to tax their own population, not Apple. Russian GLONASS is an archaic POS that they them selves have trouble working on their own, approved devices and a 25% import duty on GPS-capable devices won't fix it.

According to http://www.macnewsworld.com/rsstory/64365.html, MegaFon and VimpleCom are legitimate carriers of the iPhone in Russia.

Tim Cook mentioned sales growth in Russia as something to keep an eye on in Apple's latest conference call: http://www.apple.com/quicktime/qtv/earningsq411

The GPS chips include it (and Galileo) anyway since the frequency is close enough that you don't need any extra HW and with ASICs it's easier to just put everything it and fab a single device.

Whether the ASIC implementation works and whether the iPhone API actually uses it - is another matter

Raw precision is only part of the story. GPS + GLONASS provides a number of significant practical advantages over GPS alone, mainly due to the increased number of satellites.

Performance in urban canyons or heavily wooded areas is greatly improved, as there's a much better chance of the receiver seeing enough satellites. Likewise, startup times can be much quicker.

An intelligent receiver using GLONASS only when needed could have dramatically better real-world performance with only a minor increase in battery consumption.

We're inevitably going to see GLONASS everywhere due to the Russian import duty, but it's a real boon to most ordinary GNSS users, for whom the main limit on practical precision is being able to see enough satellites. This is likely to become increasingly important as the existing GPS satellites reach end-of-life.

though off-topic and a bit of an aside, at university (I did civ eng) I learnt that it was possible to get cm-levels of accuracy out of GPS through clever arrangement of multiple sensors (for e.g. measuring the movement of buildings due to subsidence, etc.)

In fact I just found a wikipedia article all about it - http://en.wikipedia.org/wiki/GPS_augmentation :)

In mid 90s I worked for a company that developed combined GPS/GLONASS receiver and aimed specifically at high-precision measurement market. With clever math modeling and by letting a receiver rest on one spot for longer periods of time, e.g. 30 min, they could get the precision down to 3 mm (millimeters). And that's without A-GPS, purely through oversampling of satellite data.

It's funny you mention that, as I swear the professor said it could be done to accuracy of mm, but after reviewing the wikipedia articles I had concluded I misremembered. Perhaps not...

You can, but you need to need to analyze the phase of the GPS signals at the analog level (not the the code phase)

So the carrier phase is derived from the Cs/Rb clocks on the satellites? That's pretty neat, if so.

That company was Javad, no?

It was Javad's, but before Javad, the company.

They use this on automated farms. At known co-ordinates, they stick a GPS receiver in the ground. The difference between the known co-ordinates and what that GPS receiver sees is then continuously transmitted to the vehicles on the farm.

It's called rtk correction: http://en.wikipedia.org/wiki/Real_Time_Kinematic

There's an opensource rtklib which supports a handful of inexpensive receivers: http://gpspp.sakura.ne.jp/rtklib/rtklib.htm

Why don't cell towers do this?

Cell towers kind of do by providing an accurate time and other features in AGPS.

Not really. AGPS improves the time it takes to lock on to the GPS signal, but it doesn't improve accuracy.

You need a receiver in the phone that can phase track, which costs 50x as much as the simple GPS chip they fit.

You also need either a lot of storage locally or a fast data link between the base station and the rover - phones actually have both of these but they cost money

You need a receiver in the phone that can phase track, which costs 50x as much as the simple GPS chip they fit.

So, $0.50 instead of $0.01? :)

I think Phase tracking GPS chipsets are still pricey - outside surveying applicatiosn there isn't much demand.

Yes, until the property market crashed in the US I was a land surveyor. This is the same principle used in RTK/GPS (real-time kinematic) measurement. A "zephyr" GPS antenna is placed on a coordinate "base" usually 2m above it on a tripod, which then continually updates and measures the difference in distance between it and the "rover" antenna carried by the surveyor, as well as the satellite network.

I used to have an updating map on the web of all the GLONASS satellites the Russian government had put into orbit, but it was taken down.

Thats very interesting. Do you have any information on how much these farms are automated?

Autosteer is quickly gaining in popularity. From looking at the farm equipment in my area, I'd guess that 40% have GPS systems on their tractors/combines based on observing the antennas on the equipment. Precision farming is becoming important to increasing yields and reducing costs.

Right now, Kinze are heavily touting their autonomous system for grain carts: http://www.youtube.com/watch?v=MhA5aIw7xNk

John Deere and CNH have similar systems in actual production, but still require an operator to be present for safety reasons. And safety seems to be the big limiting factor. While the technology is basically there, nobody wants to remove the operator just yet. Though in this video, someone disabled the safety sensors to go fully autonomous: http://www.youtube.com/watch?v=zU4liQvrcm4

Most farmers are pulling their RTK data over wireless/mobile internet connections, rather than having their own base station. I don't know of anyone who maintains their own station. A few companies have placed base stations around various places in the countryside and will happily sell you their data.

maybe interesting:

In Europe (and of course in Germany) the federal governments offer realtime gps correction data via IP or 3G (radio was shut down a few years ago).

They offer a precision of 1cm-2cm used together with GPS/GLONASS

PDF: http://www.sapos.de/pdf/Flyer/2004Flyer_e.pdf

(Sorry, German language only:)

http://www.sapos.de/ http://sapos.bayern.de/service.php

In 1995 I worked for a GPS developer, Avcan Avionics, and we built software that would allow tundra devices in the Arctic to travel between a supply depot and remote mines in 100% whiteout conditions, navigating only by a path that had been loaded into GPS. Our variant of this process was called "Differential GPS" - basically leaving a base station in single location for 24+ Hours where it would eliminate the S/A error which was presumably random.

We successfully navigated a car down a twisting road, about 10' wide, without looking out the window, so it was a useful mechanism. It became less important, of course, once the US Government removed the S/A error from the GPS signal and the commercial use of GPS took off.

I'd imagine that if you just left a GPS receiver in the same place for long enough, the movement of the GPS satellites would provide some decent dithering for the signal, letting you get fairly decent accuracy just by averaging a lot of measurements. I'm sure there are other methods that work more quickly, though.

I know carrier phase differential GPS is used on satellites to get amazing accuracy. I'm not quite sure what the implications are on mobile phones.

At mine sites, we used http://en.wikipedia.org/wiki/Differential_GPS

I don't understand from your table how the combined use of GPS+GLONASS would have a worse 'best case', 2.00 for GPS alone vs 2.37 for GPS+GLONASS.

There probably is some general statistical explanation of why you can get worse accuracy from combining multiple sources than each source by itself, but it seems irrelevant.

In actual receiver implementation you really get worse accuracy for simpler reason: most of the hardware resources (ie. receiver channels) are same between GPS and GLONASS, so using GLONASS ties up resources for GPS and vice versa.

I don't think this is true. Most of the cheap GPS chips I see on sparkfun can track 60 satellites at once, and you're not going to be seeing that many at one time anyway.

I would imagine for the same reason as wearing two watches doesn't improve your knowledge of the current time. You're more likely to be within the largest margin of error (i.e., the latest time on either of the watches is probably no later than the current time), but what can you say about the earliest time? Since it's equally likely to be correct on either watch, the only thing you can do is average it. But in the case that one of the watches is actually correct, you've just introduced error.

So, if I understand you correctly, the average error decreases but the best error actually increases?

I don't know the explanation, but I know this comes up in biometrics a lot. E.g. using fingerprints + facial geometry to identify someone is much less accurate than using either one or the other. Bruce Schneier has written about this in his book Beyond Fear, and I'm sure there are explanations online somewhere.

If you expect that more information will decrease performance, your algorithms are seriously borked.

http://yudkowsky.net/rational/bayes

Also from the same page:

    Some modern receivers are able to use both GLONASS and GPS
    satellites together, providing greatly improved coverage in
    urban canyons and giving a very fast time to fix due to over
    50 satellites being available. In indoor, urban canyon or
    mountainous areas, accuracy can be greatly improved over 
    using GPS alone.

Is there any reason, other than the higher number of sources, why GPS+GLONASS is better than just GPS in these situation? I.e. Is GLONASS better in these sorts of environments?

In situations like the "urban canyon", where the problem is that only a small slice of sky is visible, I would think that the advantages of using both systems would only be marginal.

The most accurate GPS fix is had when the satellites you're seeing are widely separated in the sky. In the urban canyon, the problem isn't a lack of satellites, but the fact that you can only see those that happen to be in a narrow slice. And narrow slices are bad for constraining the solution to the equations.

The way I think about it is each satellite being used in the solution is like a rope tied to the satellite (where stretchiness of the rope is analogous to measurement error). Your GPS receiver can calculate your location to within the accuracy of how far you can move while keeping all ropes taut.

So if you had a fix from two satellites that are smack next to each other, you really don't have a fix any better than a single sat. If you've got two ropes at 180-degrees opposite each other (i.e., at either end of the street in the urban canyon), you've still got ambiguity due to your potential for lateral movement. For a good fix, you've got to bring in another rope at a wide angle from those -- but the tall buildings interfere with getting that rope, even if there are lots of satellites up there that could be offering it.

Yeah, but if there are more useful satellites in the sky, there's a greater chance of the visible satellites having a wider separation. Even in the 'urban canyon', you still frequently have partial sky visibility in some non-vertical directions. And there will be occasions when a GLONASS satellite will be better placed than a GPS satellite.

To use a reductio ad absurdum argument, imagine a satellite positioning network with a thousand nodes. Wouldn't it have a greater chance of providing a useful fix in this environment: http://g.co/maps/sa8zk ?

It's all about (visible) satellite count

I'm wondering whether it actually matters to people whether their GPS device has 2, 4 or 7m accuracy. Anything with lower resolution than about 5 cm is going to be largely useless for robotic driving, etc, particularly anything worse than 2 m res - so really (and not meaning to be rude to you and your research efforts) - isn't the different resolution irrelevant, and isn't the only reason GPS users should be happy about GLONASS integration is redundancy (especially in the rare military event that the US decides to switch public GPS off?)

Computer driving is more about vision systems than GPS. Maps certainly aren't accurate into the centimeter range in any event.

Sailors find that their charts are some times off by kilometers. I've personally seen a large, erroneous discontinuity in the western coast of St. Lucia while using a GPS chart plotter for navigation. Chart data is copied forward from older charts, many of which were made in the 1800s, those guys were good with a sextant, chronometer, and a pen, but when near land you need your eyeballs.

If your maps were wrong you could start a war :-) [1]

[1] http://abcnews.go.com/blogs/politics/2010/11/google-nearly-s...

Wow. Shouldn't it be strate forward to correct such large errors by using public available satellite images?

Think of augmented reality apps, 2 vs 7 meters can be a big difference here if you are close to the "to be augmented" target.

Also for OpenStreetMap this would be great. This would get the accuracy down to "error less wide than a street" levels on customer devices.

It means more satellites in the constellation(s), and in circumstances where that makes the difference between getting a GPS/GLONASS fix and having to rely on cell tower/WiFi positioning then the difference in accuracy would be very significant.

Isn't that point confirming what I said, albeit confusing redundancy (availability of other systems as backup when GPS fails) with accuracy?

I'm not talking about the entire system failing and falling over to the other, I'm talking about the probability that from any given point you can see enough satellites to get a fix. More satellites is better.

I would love it if I could use "find my iPhone" inside my house without having to resort to the audible alert when other people are sleeping. So, yes, better resolution would be wonderful. Edit: and to address the obvious objection, no, the current resolution is not good enough for this. With the current accuracy, I'm not able to locate my phone within my house yet (iPhone 4, and no canyon issues here).

Anyone tested iPhone navigation in a gps blackout area (e.g. Near an air force base?). I used to live in Lompoc and GPS wouldn't work in most of the town owing to the proximity of Vandenberg AFB.

I wasn't aware there of existence GPS blackout areas anywhere, esp. in the continental United States. Do you have more information that would support your statement?

For what it's worth, GPS seems to work fine for a lot of people around Lompoc, CA: http://www.strava.com/rides-by-country/united-states/califor...

The evidence is personal experience. The blackout only affects a certain radius of Vandenberg which only covers a part of Lompoc.

So, what exactly happens there? Your device stops seeing satellites (loses lock)? Or does the reported uncertainty become on the order of hundreds of meters? Or does it simply give you confused directions when you're driving?

Searching for combinations of terms like "GPS, blackout, airforce base, lompoc" only brings up this thread. I find it hard to believe that if it were a real occurrence no one else would notice and write/complain about it.

GLONASS uses nearly the same frequency range as GPS, so the answer is no.

Tangentially related - IIRC GLONASS also happens to have an almost identical frame structure to GPS. Very peculiar fact given the context.

GPS is already better than 2m with the use of Wide Areas Augmentation Systems such as WAAS or EGNOS. This is already widely supported by chipsets.

Does anyone know if it's possible to see on an iPhone 4S if it is connected to GLONASS satellites (either through an API call or in the settings)?

GPS/GLOSNASS/GALLILEO work by sending time signals, there isn't really a "connection" to a GPS satellite. Then there is the AGPS/Wifi GPS stuff which gives you even more information. All these sources are combined mathematically to estimate spacetime coordinates.

What the iPhone API exposes is the estimated accuracy of all these sources combined.

I wonder if that's causing the odd jumping about I'm seeing on the 4S compared with the stable tracking of the 4.

GLONASS+GPS is more accurate than just GPS, so the answer to your question is no. :)

Only if your software is aware and dealing with it correctly, I'm guessing.

There isn't a GPS / Galileo / GLONASS / AGPS api in iPhone, you just get location data.

Apple generally doesn't expose APIs that require working knowledge of general relativity.

I've been having significant jumps with my 4S compared to the 4 too. It seems like a regression in GPS quality so far.

Just for accuracy? What exactly does this mean (why is it on top of HN).

Better accuracy is a big deal. For instance, it improves the quality of Augmented Reality applications on mobile phones. It could also enable new applications that people haven't thought about yet.

Samsung has added it to their WP7 devices with a patch. Probably for the same reason - the russian 25% import duty.

Not in the UK—http://www.apple.com/uk/iphone/specs.html

Are they just not advertising it in the UK though? It doesn't show up on the Australian site either, but I was pretty sure Apple was the one company that didn't do regional hardware variants.

ALSO: "Fingerprint-resistant oleophobic coating on front and back" WHAATTTTT

Related: the two first Galileo satellites (http://en.wikipedia.org/wiki/Galileo_(satellite_navigation)) will be launched tomorrow. The claimed precision of the free-for-all service is as granular as 1 meter, with the military/commercial service claiming to reach centimeter precision. I wonder when this shows up in smartphones.

Accuracy is of course important, but perhaps more important is how we build in scaleable redundancy as the number of satellites reduce. It's very frustrating when my sat nav or run keeper can't get a good signal, this inevitably leads to people dropping services or at least using them less. They key question is at what level is the service too compromised? and what's the next best alternative?

Wonder how feasible it would be to have a single GPS+GLONASS+Galileo chip that automatically feeds you the service(s) with the best signal; without murdering your battery?

http://en.wikipedia.org/wiki/GLONASS#Accuracy

There are already chips that combine the results from GPS and GLONASS. Because it moves from 31 satellites to about 50, time to get a fix is improved (since it might see 4 GPS satellites and 2 GLONASS ones more easily) and accuracy is better.

Down to 1 meter is open access, down to centimeters is commercial. I thought the open access variant would merely be on par with GPS and you would have to pay for better positioning.

I don't see a reference to GLONASS in the sub's page.

Edit: It angers me that high-ranking users abuse their priviledge to downvote. I'm making a legitimate point here; I don't see "GLONASS" or any reference to alternative GPS in that page.

On Reddit, someone suggested that TIL links to Wikipedia should point to the specific information, not the whole article. It got >1000 upvotes. When I point out that there is no obvious link to the claim made in this submission, someone rates my comment as useless.

The first time I opened the specs page (right after this was posted), there was no mention of GLONASS for me either. Looking at google's cache atm, it's not there either ( http://goo.gl/ulMM7 ). When I refreshed the page 30 minutes ago, it had appeared, so my guess would be that this has something to do with distributed cache servers.

It's right there under "Location". You can search the page for "GLONASS" with Ctrl-F or Cmd-F depending on your OS.

Here's what I see, it's right by GPS http://cl.ly/3m0L152W3M3T1Q2V2J1L

Ctrl+F -> GLONASS. It's there.

Unfortunately I'm seeing the same thing. the first thing I did was search for GLONASS.

Assisted GPS

Digital compass

Wi-Fi

Cellular

I did a CTRL-F on GLONASS and no match is found. I said I did this in my edit... I'd take a screenshot, but I'm at work.

Where do you live? It's probably GPS-only for your region.

No sign of it on the Canadian website. Now you've got me curious if the Canadian version of the phone has it disabled or something, because its presence would really help me with an app I am working on.

Assisted GPS and GLONASS

That's what I'm seeing from Europe.

Sorry, there was no way to have a link to the related section on this page.

Reply on news.ycombinator.com