A project (http://dronelab.missouri.edu/projects/eon-underwater-remotely-opera...) that we did a few years ago was building a ROV - we used an ethernet cable (50') that was zip-tied to a lightweight steel cable. This allowed us to use ethernet (100mbit/1gbit) to communicate with onboard Pico ITX linux box, and also to have a safety tether. Basically, we had enough bandwidth to stream HD video to the surface, and could also pull the ROV to the surface if it malfunctioned.
The goal of the project was to take 3D underwater footage to aid in underwater rescue situations - the thought being a test scenario would be a sunken vehicle (conventiently, there's one at the bottom of a local quarry pool).
Looks great !
A tech question :
I noticed that the motors are submerged in the water.
The water doesn't short the coils inside them ?
apparently not... there's quite a few videos/forum postings about people using brushless motors in water. Have to keep in mind though that the water isn't the most friendly to motors. Might need to oil/regrease bearings, and make sure the motors are thoroughly washed out after use. We tested in a swimming pool initially with chlorine, and that was pretty corrosive in the short time spent in the water.
For ESCs, we used marine/model boat ESCs - they were kept completely outside of the model in the water, so the electronics had to be water tight (but we didn't test too deep to worry about pressure).
Thanks a lot
Getting smarter every day :)
This is the general idea we used for my senior project, CUBOAT. Its pretty challenging and we only achieved a low bandwidth since the frequency is low and the noise is significant. If anyone is seriously considering a project on underwater communication I would be happy to help.
A different approach to all this is your underwater vehicle, tethered to a surface vehicle (boat) which follows it via the likes of gps and synchronous movements, and uses whatever above-water radios you please.