core: services: ping: minimise port duplication races

[ES-Alexander]

Hopefully fixes #1200 (or at least significantly reduces the likelihood of it).

Details:

• ⚠️ Untested
  • should be testable at esalexander/blueos-core, :ping-fix-attempt (in the BlueOS Version Chooser)
• Uses the approach I suggested here
• It would be fully robust to also/instead regularly check for duplicated ports in the drivers (and restart them as necessary), but that seems excessive to me, and more involved to program and test
• Ideally this code should also stop treating a port as being connected to once the connection is established, but that seems like extra effort for minimal gain (as far as I'm aware it's not important for the ports to be consistently the same, so as long as unique ones get chosen that aren't crazily far from the base port then it should be fine)
• This code also has a bit of duplication in the handling of ping1d vs ping360, but I didn't want to deal with the type checker, and reducing the duplication may end up with slightly more code anyway

[Williangalvani]

looks good and tested by Geoff here:
https://discuss.bluerobotics.com/t/ping-sonar-altimeter-udp-port-overlap-in-blue-os/14506/9?u=williangalvani