Team-BHP (https://www.team-bhp.com/forum/)

-   In-Car Entertainment (https://www.team-bhp.com/forum/car-entertainment/)

-   -   GPS epoch rollover (https://www.team-bhp.com/forum/car-entertainment/208356-gps-epoch-rollover.html)

Mods, Not sure where to post this. Please move this to an appropriate category. Thanks in advance.

Hope this information is not too late. GPS rollover happened on April 6th. This has the potential to affect the devices we use. GPS has been embedded in various devices that we use off late like Phones, Cameras, Car Navigation Units, Exnternal GPS chips. This event is akin to the Y2K event we witnessed in, well, Y2K.

I did update my firmware of external GPS chip last week. I am not sure it contained a "fix" per se, but it sure is working fine.

Please go through the links given below to more understand the nature of issue.

• https://arstechnica.com/information-...-side-effects/

• https://techcrunch.com/2019/04/06/gp...ght-get-wacky/

And, if you think that all is well in the world, read below.

• https://simpleflying.com/boeing-787-china-grounding/

With the recent MCAS issues, and GPS rollover issues, Boeing is really not inspiring confidence these days.

Did you face any issues off late that could be attributed to this? Is everything functioning ok? Sound off below.

Oh, I am not worried about my mobile phone, I am sure the software update would have been released to take care of it.

But my in-dash GPS device from 2014, I don't think it can handle this. Need to verify whether it will remain accurate.

We had a tough time out at sea where almost all navigational equipment are connected directly or indirectly to a GPS/DGPS receiver.

Preparation for worse possible effect / emergency drills etc carried out to prepare ourselves, including verification of each GPS make/model, if it is effected or not, contacting makers on same by our shore-offices. Various makers came up with their own solutions to overcome the issue (if make/model is effected). Also as per various GPS manufacturers, this phase is going to last till August 2019.

In one word I can say, if the subject GPS built after 2010, there will not be any issue (they are designed to tackle this).

Let me allow to be more specific and explain in layman’s language.

The GPS system is synchronized to a different time scale than Coordinated Universal Time (UTC) and calculates time by counting the number of weeks up to 1,023rd week. After week 1,023, the week number count is reset to zero. This is called "GPS week number rollover". Subsequently, the erroneous date information on the GPS receiver clock may result in instability in positioning, causing position errors, including failure of GPS position acquisition, etc.

When GPS week number rollover occurs, our GPS Navigator / Sensors may cause the following problems:
1. Date displayed rollovers to 19.6 years back.
2. Due to GPS rollover, equipment positioning operation, date/time display may become unstable or may cause equipment failure.

So just check out your devices, I think we all will be safe.:thumbs up

Cheers,
Sail

I have a 2014 Garmin purchased in the US; got an email from them two weeks back saying there is a firmware update available for download that I should install before using the device.

Most GPS (old or new) will tackle the roll-over seamlessly if it is ON *during* the transition, i.e. if it was switched on few mins before Apr 7 00:00UTC, 2019, provided it has a memory to store the last valid time.

Issue is with the units that are being switched on after the event. Again, it depends on the availability of battery backed up real-time clocks (RTC) in these units. If there are no RTCs, the date will roll back to August 1999, that is the starting of the previous epoch. A firmware update in this case can help, where it stores a hard coded default date that is past the Apr 7 2019.

Any units that has alternate source of time (phone, GSM network) will not have the issue. Also newer units are capable of using GLONASS satellites, that gives the full calendar year on its transmitted data, thereby solving the uncertainty in the GPS year.

The position accuracy of the GPS will not have any impact even if it is affected by this issue. Only the calendar time output will be wrong. Of course, any other application that depends on the GPS (calendar) time will be impacted.

Its much complicated in avionics, as a vast number of systems depend on the GPS time. The Boeing 787 had advisory circulars on how to handle the situation and majority of the planes did not have the issue. Only a handful that was grounded and turned off during transition are affected.

Surprising to see just a few 787s affected and then also from the same carrier.

They must have been equipped with some sort of odd piece of kit that would not be part of the mainstream maintenance bulletins and routines.

You mention Glonass, but how does a mobile phone decide which system to use? Is it user setting or does it switch automatically?

Jeroen

At least in iPhones, the hardware/software combination decides the location based on many factors: GPS, mobile telephony transmitters and WLAN. Usually the GPS chip will get its bearings from the 4 satellites of the GPS if available and check Glonass for supplanting the location information if some GPS satellites are not in view.
Glonass anyway is not that accurate in all areas except in far north or far south.

I got a GPS specific update on my Lenovo mobile phone sometime last week. This phone is more than 3 years old and doesn't receive any firmware updates since Android 6.

The update mentioned that it was specific to GPS and cautioned GPS might not work properly if update is not installed.

GPS needs 3 satellites to calculate a position, fourth satellite will make it more accurate. Even with 3 satellites it is likely to more accurate than any position you can get out of Mobile Network.

I am not quite sure what you mean with GPS and Glonass are not compatible I thought? Totally independent systems using their own standards . Are you saying that a location would be derived from 3 GPS and 1 Glonass satellite?

I would think the iPhone uses whichever system can get it the most accurate location. So if you find yourself in a location where GPS shows 3 and Glonass 4 satellites it would use Glonass.

I would not think it combines the two systems into one reading. It is one or the other, but then again I might be wrong.

Jeroen

Yes, both GPS and GLONASS are independent systems, but satellites can be mixed to derive a position, once the difference in the reference time between the two is known. This is typically employed in the GPS chips of new phones.

One reason for the adoption is that the usage of GLONASS system is being mandated by Russia in various sectors (usage in Russian territory, by Russian registered operators), including aviation, emergency responders, etc. So the OEMs are packing multi-constellation capabilities in the same chip/unit.

I hear what you say, but multi-constellation to my understanding means it will use a GPS constellation or a Glonass constellation, what you are suggestion is a mixed constellation. Which is something entirely different. That means deep integration of the two systems way beyond settling the clock reference.

Do you happen to have any more technical references to how they achieve this?

Jeroen

I may have oversimplified it, but it is being done. There are a lot of papers available online. Just one of them:
http://citeseerx.ist.psu.edu/viewdoc...=rep1&type=pdf

Thanks.
This paper shows that it can be done, not that it is being done. I am interested so I have searched around a bit. I have found quite a bit, but I haven't seen anything yet that suggest that there are chipsets available and being used to provide proper integrated positioning. Rather than switching between two different constellations.

Let's what else comes up. It is certainly an interesting development.

Jeroen

It is already being done. Infact, that technology sort of existed even before multiple systems existed. The reason, is that part of using satellites for a fix is done in "software" or firmware.
The purpose of the silicon and antenna is just to receive the data which then goes to the processor.
So with different freq bands, you may require a more complicated antenna + silicon to handle and decode different bands. But once that is done, its handed over to software

As of today, the quad band receivers are complicated
https://www.gpsworld.com/quad-conste...alileo-beidou/

After 2020, the silicon will be simplified.

Now coming to what we have now, the software already can get fix using multiple satellites of different constellations
so you may have 3 USA and 2 Russia satellites, and get a fix.
A more technical answer on stack exchange
https://electronics.stackexchange.co...ioning-modules

It also references papers.

Such systems are called GNSS
One such receiver (Dual band though) is Titan X1
https://www.mouser.com/ds/2/813/Glob...et-1223989.pdf

There is also an urban canyon test for GPS only, GLONASS only and combined.
https://www.electronicdesign.com/tes...ict-gpsglonass

thanks, some very interesting reading. I wonder how this would work for speed readings, as the underlying technique on GPS tends to be a doppler measurement on some of the signals. Not sure how GLONASS or the others measures/calculates speed.

With different orbits, different reference clocks etc, it will need quite some corrections to get a speed using satellites from different constellations.

Jeroen