Hi. Relatively new to the field (survey student). Question about RTK corrections...
Short version:
What instrument generates/computes corrections for an RTK rover: the base or the rover?
Long version:
I have read that a base station "sends corrections" and that a rover "receives corrections". To me, this suggests that the base knows where the rover is (via some transmission from the rover), generates "personalized" data for that rover, then transmits that data to the rover. Seems like these corrections would be valid for only one rover.
I'm having trouble buying that as I don't think the rover transmits anything, and what about multiple rovers off one base? I suspect the following:
The base transmits raw data about its position as a passive signal for anyone listening, then a rover will generate its own corrections and apply them. In this scenario, the base is not literally sending corrections. But I keep seeing that phrase so I'm kinda wondering...
Can anyone shed light on this? Thanks.
-Ben
When a base station is started it is seeded with correct coordinates for it's location. As that base receives GPS data, it compares the coordinate compute using GPS to the seeded coordinates, and then broadcasts the difference (corrections) out to the world.
Others will go into this deeper and better than I, but the assumption is that the rover won't be too far away from the base, so the corrections at the base are equally valid at the rover. All the rovers get the same corrections, those of the base.
> As that base receives GPS data, it compares the coordinate compute using GPS to the seeded coordinates, and then broadcasts the difference (corrections) out to the world.
To amplify this a little, the base broadcasts its position and the range data from the SVs that it sees. The rover combines that data with the SV data its antenna receives, solves the vector to the base, and calculates its position by using the base position and the vector. This is done at each epoch; typically once per second, but in some cases 5 times a second or more.
The rover can solve the vectors this quickly because it maintains the integer ambiguity solution from one epoch to the next, as long as it maintains lock on the requisite number of SVs. When it loses lock it has to reinitialize (resolve the integer ambiguity again).
As for the rover sending info back to the base, that's something that happens with VRS systems. The network receives an autonomous rover position and uses it to create a virtual base very near the rover in order to reduce the systematic error associated with longer vectors.
All of the above is based upon a rather hazy understanding of the way all this works, so if I've got parts (or all!) of it wrong, I trust someone will set things straight.
Just this small clarification, it isn't the vector that gets solved, it is the position of the rover that gets solved, but we like dealing with vectors, ie, azimuth, distance, and delta height, so that is how they solution gets displayed for our use. It is the inverse of the coordinate differences along with the solution errors.
That is known as DGPS and applies to code measuring receivers.
Bingo. The data is then stored in a format we are comfortable with. Hopefully delta xyz with q info...
Thanks for the replies. I wasn't making the connection to the seeded coordinates. I guess I was thinking of the corrections as being merely the deltas between the base and rover. Enlightening.
-Ben