The construction consists two raspberry pis, model 3 with an Aldebaran modul from Dr.Fasching and each a Tallysman tw-2410 antenna.The goal is to realize a system with a moving-base and a rover modul with RTKLIB. The Base should get its own position automatically by GPS without entering any position coordinates during or before the moving process. The rover should get its own Data from its own GPS modul. Base and Rover are connected by WiFi. The goal is to print out the relative distance in real time.

I established the Base modul as an Wifi access point (AP) und connected the rover as a client to the base. Base and Rover swapping their data by TCP.

At this state i have the normal fixed-base, moving-rover solution. It´s also not clear why some satellites are chosen from the system/rtklib and why others are not, although i receive good signals. The data transfer is realized in nvs binr. Take a look at the sketch in this file-ordner.

To set up my raspberry pi3´s i followed the instructions of:

1. Dr.Fasching: http://drfasching.com/products/gnss/raspignss/installation.html and

2. custom-build-robots: https://custom-build-robots.com/raspberry-pi-roboter/praezise-gps-positionierung-rtklib-navigation/7958#comment-1376

On the custom-build-robots the instructor changed the rc.local script on the raspberry pi(rpis) which i did not. Instead of changing the rc.local script I wrote: 1.nmeafile.nvs 2.binrfile.nvs so you easily can switch between those scripts without having a bunch of commands to tap.

I also build the rtklib 2.4.3 on both of my rpis. Like i mentioned before at the moment i still stuck with the fixed-base, moving-roving solution, trying to realize the moving-base, moving-rover solution. I tried some settings from https://rtklibexplorer.wordpress.com/. I know that the title DGPS is a bit misleading for what my project actually is, cause i am not trying to use any reference stations.