Hi all,
I'm wondering what the drawbacks are for using TCP vs UDP for the connection between the flight controller (via mavlink-router) and GCS over a semi-reliable connection (4G)?
Is there any difference in behaviour for TCP vs UDP?
For example when a connection interruption occurs:
- In the case of UDP I assume packets will just get discarded
- and TCP will try to retransmit as long as the connection is established
I assume TCP might be annoying when sending control commands and retransmissions occur? Ie. commands "batched up" and processed when the connection is re-established. Or does the MAVlink protocol have some kind of features to deal with this?
(Still studying up on mavlink, forgive my ignorance!)
In general, I am looking for any advice or experiences on using TCP vs UDP for the mavlink connection between device and GCS.
In my specific case I'm dealing with a network that is not UDP friendly, so I wonder if I can simply switch to TCP, or if I should invest in changing the network setup.
Best regards,
Dieter Van de Walle