I have had a similar idea for quite some time, but never had the time/resources to play too much with it. In my experience it is better to get a functioning prototype of the hardware on a tether first. Things I ran into I really didn't expect. I.E. I decided to use 6 in thin wall pvc from Lowes. I cut a 1 ft section of it to test buoyancy and I filled the pipe all the way up with water while capped and it still floats too much!.
To speak to your specific wireless connection issue, I have seen guys on the internet use a small tether to a floating antenna to achieve wireless for further distances in shallower areas. Also, if you look on youtube for RC Submarines conventional 2.4ghz radios seem to be able to reach the sub. Allowing their operators smooth control almost all the way down to the bottom of the diving end of public pools. That being said, how far do you want to go?
My concept is that I want a sub with several cameras. At least one dome one on the front and one on a track sliding forward and back the whole length of the sub looking through a clear floor. I believe that using exhaust vent pipe (brazed and sealed) will give me the weight I need to avoid rocketing back to the surface. If that goes well I am going to try working on the electronics and com. systems. Most likely it'll be a simple pipe with ballasts, pressure sensor and Pi controller only. Communications will be USB Wireless adapter > Ad Hoc networking being broadcasted from my laptop. I'll calibrate the pressure sensor to be able to read depth ( Pressure of water doesnt change, just the weight experienced due to depth) and have a RPI read depth, ping my laptop and log both results. At that point let a little bit more water in the ballasts and repeat .
Using the previously mentioned method you will be able to have a log and find out what your range(at what depth does packet loss start occurring) is. Regardless of equations, experience, or even dumb luck you don't know what is going to happen until after you do it. Here are a few things I would worry about enough to test and get data on what actually happens with your specific sub.
Test different types of material, if your radio is in the hull, different materials bounce RF signals differently.
Is the water clear? After a rain storm it may leave water not so clear, potentially adding interference. Or maybe pond scum?
See how solid objects slightly in the way affect communications. If your at a local park with a pond, does the wooden pier your diving around make a huge difference?
If you are trying to push what you can do, just for the sake of doing it documentation is key. Record / Write as much info as you can to later look back on. I never noticed it until my sister pointed it out, but we can't hear trains passing through our town normally. When it storms or is very cloudy we can hear it at the same place we previously couldnt. When the clouds are there and low the sound actually bounces off them, then the ground, then back again making the sound carry a lot further. My point in this example is that routine/regular things may affect any experiment. So try to get as much info to be able to re-create it as possible. Dont overlook anything.
Sorry for the rambles
Hope it helps