Precise point positioning

[Jeroen Balduyck]

I had an interesting conversation with Gary on PPP after reading his great tutorial here https://gpsd.gitlab.io/gpsd/ppp-howto.html with regards to my plan on getting a U-Blox ZED-F9 for PPP.

Snipped (me talking):

> Now for the thing that confuses me deeply. I have done a lot of
> reading, and it appears PPP can also be delivered via satellite or
> Internet,
Answer (Gary):
> Nope. PPP is after the fact computing of position. UNtil there is
> a data center on a sat you can talk to two way, it wil have nothing to
> do with sats.

However, previous U-Blox chips (like the NEO-7P here https://www.u-blox.com/sites/default/files/products/documents/NEO-7P_ProductSummary_%28UBX-13003350%29.pdf) had PPP-processing onboard, combining it with SBAS (EGNOS, WAAS, ...) SAT-data. This is definitely one way communication and far from a datacenter. 

and https://www.navcomtech.com/en/product/globalcorrectionservice/ broadcasts should be even better (as in better predictions) and can also be one way broadcast communication.

So the info is a little conflicting? I suspect he means the computations themselves (individual results for connected receivers) cannot be send by sattelites as SATS only send the data required for computations.

Furthermore, it seems the low cost receivers seem to move away from 'internally processed' PPP in favour of RTK whilst PPP can be so useful at remote locations.

As far as I can tell PPP (in dual frequency GNSS receivers where it corrects for ionospheric errors) is almost as good as RTK apart from PPP not being able to correct tropospheric for errors (or it can, but not as good as RTK can?) to the same degree as RTK and as such it is a bit less accurate. Post computed PPP is on par or better than RTK it seems.

Can someone (Gary is most welcome to elaborate) comment on this?

Many thanks

Jeroen