Underwater communication

[maverik007]

Hi everyone,

I am looking for a communication module that would allow to underwater robots to successfully carry out full-duplex communication for coordinated tasks.

Any suggestions would be greatly appreciated. Btw, I am looking for a comm module that can be interfaced with a PIC microcontroller.

Thanks in advance for your help.

[marshallf3]

Submarines used very, very low frequency RF.

[beenthere]

Are you looking for audio? - http://en.wikipedia.org/wiki/Underwater_telephone

If not, a blue-green laser may give better range, especially if both tranceivers are going to be submerged. Otherwise things get big and power limited - http://en.wikipedia.org/wiki/Very_low_frequency

[Dyslexicbloke]

There is a list of diver com system here:-
http://ieeexplore.ieee.org/xpl/freea...number=1090687
The modern ones are almost all modulated sound carrier, something well above human hearing.
If you used light wouldnt it have to be omny directional or at least on a narow plane.
Whats the range / depth / water quality / background noise profile?
Al

[maverik007]

Originally Posted by Dyslexicbloke

There is a list of diver com system here:-
http://ieeexplore.ieee.org/xpl/freea...number=1090687
The modern ones are almost all modulated sound carrier, something well above human hearing.
If you used light wouldnt it have to be omny directional or at least on a narow plane.
Whats the range / depth / water quality / background noise profile?
Al

Interesting paper! Thanks for the link The target application is comm between two robots underwater, operating in dirty water (muddy etc) ... can't use opitical comm due to dispersion problems. I was considering finding some sort of a "tag" used to track fish, that's usually used for communicating with hydrophones floating in the water, but these send pre-determined information to the hydrophone receiver. I'd love to have control over the data that's transmitted to allow information exchange between robots (in addition to the tag-identifier that's sent by default)

[maverik007]

Originally Posted by beenthere

Are you looking for audio? - http://en.wikipedia.org/wiki/Underwater_telephone

If not, a blue-green laser may give better range, especially if both tranceivers are going to be submerged. Otherwise things get big and power limited - http://en.wikipedia.org/wiki/Very_low_frequency

Yes, I am looking for audio, as light would be infeasible in muddy environments (like a swamp)

[Dyslexicbloke]

Another thought ... have you looked at ultrasound?
I dont know anything about it but I do know it only works through a relativly dense fluid because thats what babys float arround in.

Just pondering
Al

[windoze killa]

What DB said. Ultrasonics is probably the easiest to use. It will be a pretty slow comms link but it would work. Most ultrasonics are around 40Khz. Doing some sort of digital frequency modulation would be the simplest method if you can find some that can have their output frequency changed. You could have say 38Khz for a "1" and 42Khz for a "0". Something to think about.

[maverik007]

Originally Posted by windoze killa

What DB said. Ultrasonics is probably the easiest to use. It will be a pretty slow comms link but it would work. Most ultrasonics are around 40Khz. Doing some sort of digital frequency modulation would be the simplest method if you can find some that can have their output frequency changed. You could have say 38Khz for a "1" and 42Khz for a "0". Something to think about.

That's an interesting point. I think I might be able to do something in that direction. I'll read more about it and get back with my observations and questions. Thanks

[Dyslexicbloke]

I am just guessing here but shouldnt you be able to utilise ON / OFF pulses of sound just as you would use a modulated IR system.

ThE the pulses would have to be much furtrer apart because of the time taken to reach their destination but I cant help wondering if an exiating coding system couldnt be used as a basis for your project.

Your biggest issue will be protocol selection, or design ....

Both / all your equipment needs to know its on line (Capable of talking and recieving) all the time.

I recomend you look at two line and single line transmition protocols RS485 / IC2 / Modbus..... the list go's on, but you will want a simple one with a low overhead.
IE not too much header info in the packet and variable block length.

This is because you only have a single conductor, the water, you will have to have some concept of master-slave or master-master architecture that has a flow control strategy to manage when devices Talk, Listne, or decide they recieved or transmitted an error.

Granted you will probably find that most protocols you find couldnt be implimented over a connection that is both slow and has a variable transmition delay but I expect you could take one as a model and then impliment it with a modified timing structure that suites your medium.

You may be able to impliment a protocol with more than 1 transmission line RS232 for example just very slowly, but that would intail diferent modulations for each line and complicate the detectors somewhat. That said it would be faster by some margin.

You could relativly easily design your own protocol but I recomend looking what is availlable now so you understand the challanges and potential solutions.

I find this whole concept interesting and would apriciate your sahring your progress with me if you wouldnt mind.

Al