How does differential GPS work

[sludger13]

I figured out differential GPS (DGPS) corrects incoming signal time divergencies resulting from atmosphere and malfunction of sattelites. The station knows its accurate position and accurate position of sattelites -> it calculates the distance from visible sattelites -> ideal signal travel time -> divergency from real signal travel time it is recieving.

Though I haven't found the station timing. This scheme requires very accurate time and I can not imagine this technique with time correction that common GPS receiver performs.

Also, does every DGPS station use atomic clock?

 

[ScienceGeyser]

DGPS Systems

Most DGPS systems use a reference station with a known location to calculate offset error resulting from a number of sources. This offset error should be the same for a certain distance around the reference station (roughly 200 miles). The reference station transmits the error over some other radio system and the DGPS receives this signal and applies the offset to it's own readings. Actual time from the NIST (or other) atomic clock is not necessary in this system.

 

[sludger13]

Could you be more specific? If atomic clock accuracy is not necessary, then the station must correct its clock bias, right?

To perform it, at least four visible sattelites have to be healthy, in order to get an intersection (somewhere) to compare with its real known position and position of (unhealthy) sattelites. If every visible sattelite is unhealthy, the station wouldn't know as it doesn't know the accurate time or (at least) running of its time divergency. Strange situation...

 

[davenn]

BTW its satellites NOT sattelites

The "atomic" clocks are on the satellites
define what you mean by unhealthy ?

you do realize there are a LOT of satellites up there ?
between the GPS and the GLONASS sets there are some 25 +

I'm a Trimble GPS tech andwork on GPS equip. on a daily basis and I never see less than 10 satellites
more commonly, anywhere between 14 and 20 at any given time
The timing primarily comes from the satellites

Here's a link to a small tutorial from Trimble on DGPS
https://www.trimble.com/gps_tutorial/dgps-how.aspx

Here's another DGPS explanation for another Trimble tech...
http://www.esri.com/news/arcuser/0103/differential1of2.html

he can explain it better than me haha

Dave

 

[ScienceGeyser]

DGPS Systems

The DGPS reference station doesn't need to have an accurate clock since it can derive it's time from the GPS data stream. The clock data don't need to be part of the differential offset data. It simply uses the the standard GPS methods to calculate it's position and then compares the calculated position to the known position which produces an offset value. This offset value can be used by other mobile GPS systems in the area to correct their offset error which should be the same given that they are receiving an identical GPS data stream.

The health data from the reference station simply indicates whether or not the station is operating within it's normal parameters. If there is some sort of data error and the station somehow is unable to correctly supply all of the data necessary for an accurate offset correction, it will report that it is unhealthy so that the mobile GPS system will not apply some erroneous offset signal. The satellites are also sending health data and the health of the satellite constellation is also reported by the reference station in some cases.

Having said all of that, there are other differential correction schemes. This is the basic function of the WGS 84 and NAD 83 systems. RTCM SC-104 may operate somewhat differently but is should not be vastly different. I have little to no information about any other methods.

 

[sludger13]

define what you mean by unhealthy

That is the satellite which trajectory, timing or outcoming data code varies from expected condition.

you do realize there are a LOT of satellites up there

That is the only solution, if you are so sure about the quartz clock bias correction. Potentially every visible satellite can malfunction, however how possible is that to happen...
The links are fine, but there is nothing about its timing...

The DGPS reference station doesn't need to have an accurate clock since it can derive it's time from the GPS data stream.

It's the point I'm speculating about. If no satellite works properly, the station might set its time wrong from data stream -> is not watching over satellites function correctly anymore.
The atmosphere signal delay is similar for (at least some) satellites -> the calculated position varies, but calculated time is accurate.

Those troubles would come off with station accurate timing (with no clock bias correction) :)