Modulation/Signaling scheme for transfering information over an acoustic channel

On 11/9/2011 6:49 AM, John McDermick wrote:
> Which modulation scheme is most suitable for transmitting information
> (a bitstream) over an acoustic channel? i.e. a loudspeaker outputs the
> modulated signal and a cell phone microphone picks up the acoustic
> signal and extracts the information from the modulated signal. I'm
> looking  for a modulation/signaling scheme which is robust with
> respect to ambient noise in the acoustic channel.
>

As you can see, this kind of broadly defined situation conjures up all 
sorts of assumptions about what it is that you must mean by it!

It would be helpful to narrow it down:

1) In air or in water or in some other medium?  It seems that someone 
assumed "in air" but I didn't see where you said that exactly.

2) What is the order of the bit rate?

3) Over what distance?

All of these are the "system" parameters that determine whether there's 
likely a feasible solution and would set aside the silly examples that 
just don't fit the objective.

You've already been given some interesting examples:
- acoustic modems .... in air, 300 baud?, short distance.  I'd include 
handheld DTMF boxes as a simple unilateral version.
- downhole telemetry which may or may not be "in air"....
- tom toms

Now, I don't know what a "relevant" acoustic process might be.  Would it 
have to necessarily be a binary or PCM sort of implementation to be 
relevant?  Otherwise:

- PA systems ... in air, 5kbps?, medium range
- underwater telephone ... in water, 5kbps?, medium range (e.g. a few 
100 meters)


I don't think it's proper to call a flashlight or an active sonar a 
"communication system" unless there's a receiver at the other end.
So, you might tell the forest from the trees with a flashlight but you 
don't "communicate" in a sense unless there's someone to see the 
flashes.  Nonetheless, many of the same system concepts apply.
I also wouldn't include acoustic positioning systems for the same reason 
becuase the simple ones only say "here I am ... beep beep"

History is filled with cases where technology from one application is 
suggested for another application and falls flat on its face.  I'm sure 
there are plenty of opposite examples too.  The usual problem is that 
the medium is different.  Some examples:
- radar often works with multiple pulses closely spaced in order to get 
range rate.  Sonar doesn't because of reverberation and rather relies on 
Doppler. Doppler in some radars was problematic if one were to turn the 
table.  They look similar on the surface but just aren't *that* similar! 
  But the idea of a RAKE receiver was proposed for sonar because of 
varying multipath.

Fred

On 9 Nov, 22:04, Fred Marshall <fmarshallxremove_th...@acm.org> wrote:

> Now, I don't know what a "relevant" acoustic process might be. &#4294967295;Would it
> have to necessarily be a binary or PCM sort of implementation to be
> relevant?

I think I used that term, so I'll elaborate:

I consider any acoustic comm system as 'relevant' iff

- it works over some distance
- has a certain non-trivial bandwidth / symbol rate
- there are few or no practical alternatives
  to acoustic comms for the stated problem
- the system relies on electronic transducers and
  recievers for modulation / demodulation of the
  transmitted waveform, as opposed relying on the
  human body producing (possibly with tools) or
  recieving (by hearing) the acoustic carrier
  waveform to be transmitted over the channel.

Well, you get the idea.

Rune

On 11/9/2011 12:31 PM, Jerry Avins wrote:
> On 11/9/2011 1:31 PM, Vladimir Vassilevsky wrote:
>>
>>
>> Vladimir Vassilevsky wrote:
>>
>>>
>>>
>>> Tim Wescott wrote:
>>>
>>>
>>>> Hand-held DTMF generators, from back when rotary dial phones were
>>>> still commonplace, but voice-response systems were starting to take
>>>> over the WORLD!!!
>>>
>>>
>>> The ~200 bit/s is a great accomplishment, especially if you could get
>>> the DTMF to work reliably from one side of a room to the other.
>>
>> Oops; it is ~50 b/s actually
>
> I'm looking at an old 300 Baud acoustic modem built into a TI
> thermal-printer terminal.
>
> Jerry

Jerry,

Heh.  I had one of those in my office in Crystal City, VA.  I wrangled 
ARPA to get me an account so I could hook up into the ARPANET.  Around 
1983.  That's when there was something like text-based "email" between 
the few users..... well, I don't know what "few" means but that was 
nearly 30 years ago and things have change just a bit....

Fred

On 11/9/2011 12:28 PM, Rune Allnor wrote:
> On 9 Nov, 21:26, Jerry Avins<j...@ieee.org>  wrote:
>> On 11/9/2011 12:03 PM, Rune Allnor wrote:
>>
>>> On 9 Nov, 17:48, John McDermick<johnthedsp...@gmail.com>    wrote:
>>>> These papers talks about transmitting information over an acoustic
>>>> channel:
>>
>>>> http://www.cmlab.csie.ntu.edu.tw/~ipr/mmsec2007/data/homework/hw2/Int...
>>
>>>> http://www.apl.utoronto.ca/publication/i/nevena_tmm06.pdf
>>
>>> People have been *talking* and *writing* about such
>>> stuff for decades. You would be hard pressed to
>>> come up with examples of systems that actually *work*.
>>
>> How do you classify acoustic modems?
>
> As an acoustic interface to an EM comm channel.
>
> Rune

Reminds me somewhat of the term: "external locus of control"....  :-)

I think the "acoustic interface" is an acoustic communication channel 
pure and simple. What else can it be?

Fred

On Wed, 09 Nov 2011 18:44:34 +0000, glen herrmannsfeldt wrote:

> Tim Wescott <tim@seemywebsite.com> wrote:
> 
> (snip, and previous snip on acoustic data transmission)
> 
>> Hand-held DTMF generators, from back when rotary dial phones were still
>> commonplace, but voice-response systems were starting to take over the
>> WORLD!!!
> 
>> Hold gizmo up to mouthpiece, punch buttons -- joy ensues (well, as much
>> joy as can ensue when you're in Voice Mail Hell).
> 
> 
>> Old-style modems, with the handset holder thingie.  Put handset in
>> holder, joy ensues (well, as much -- oh, never mind).
> 
> But in both of those there is very close, maybe even a rubber seal,
> coupling between the two.  The OP wants to go through open air.

Oh, details.

> Maybe the OP should consider closer coupling, with the receiver as close
> as possible to the transmitter.  (Though likely no rubber seal.)
> 
> Otherwise, to keep noise away it will need to be at a very high level,
> which will not make the user or his/her neighbors happy. (and might burn
> out the loudspeakers.)
> 
> With data rate pretty much proportional to bandwidth, it will be the
> tweeter side that is most useful, the most prone to burn out at high
> power levels, and most irritating to users at high (or even medium)
> power levels.

The OP could state just how "open air" it needs to be?  I suspect you 
could do a lot better if the user was expected to hold the phone up to 
the speaker than if it had to work from across the room.

-- 
www.wescottdesign.com

>
> The OP could state just how "open air" it needs to be?  I suspect you
> could do a lot better if the user was expected to hold the phone up to
> the speaker than if it had to work from across the room.


speaking of difficult channels

it would be really rewarding for someone to develop an emergency comm system
for miners trapped underground...

Mark

On Wed, 9 Nov 2011 18:49:19 -0500, "MarkK" <makolber@yahoo.com> wrote:

>
>>
>> The OP could state just how "open air" it needs to be?  I suspect you
>> could do a lot better if the user was expected to hold the phone up to
>> the speaker than if it had to work from across the room.
>
>
>speaking of difficult channels
>
>it would be really rewarding for someone to develop an emergency comm system
>for miners trapped underground...
>
>Mark

There are standards for such things, and products that meet the
standards.   I don't know how well any of them work in practice, but
some seem to be pretty useful.

We did some consulting for these guys several years ago for some early
work on a magnetically-coupled system:

http://www.kuttaradios.com/

The idea is to use magnetic coupling to existing infrastructure,
pipes, rail, cables, whatever.  Naturally, if such infrastructure
isn't in place or gets disconnected in an event, it won't be useful.

There are other approaches as well, some using traditional radio
techniques, but it's a difficult proposition to make something that
will work in all cases.   The through-the-ground stuff is probably not
practical, either.

Over the years and especially lately I've spent a fair amount of time
in an extensive tunnel network in a silver mine.  Recently I took some
FRS radios in there with me, keeping my fingers crossed that the
tunnels would provide some degree of waveguide or power concentration
effect, but the range of the radios was about the same range as just
shouting at each other, so it didn't work well at all.   I have some
ideas of how to make that better, especially around corners, but
haven't had a chance to try anything yet.

It is a difficult and interesting problem.   I'm hoping to use the
access to the tunnels for some experiments, but won't be back in there
until next year.


Eric Jacobsen
Anchor Hill Communications
www.anchorhill.com

On 10 Nov, 00:00, Fred Marshall <fmarshallxremove_th...@acm.org>
wrote:
> On 11/9/2011 12:28 PM, Rune Allnor wrote:

> > On 9 Nov, 21:26, Jerry Avins<j...@ieee.org> &#4294967295;wrote:

> >> How do you classify acoustic modems?
>
> > As an acoustic interface to an EM comm channel.
>
> > Rune
>
> Reminds me somewhat of the term: "external locus of control".... &#4294967295;:-)
>
> I think the "acoustic interface" is an acoustic communication channel
> pure and simple. What else can it be?

An minor acoustic component in an otherwise totally EM system?
In the EM case the modem would hardly be considered a major
part of the propagation channel, but rather a component in
the interface between the source of the signal, and
the waveform that propagates as EM waves from the antenna.

Rune

MarkK wrote:

>>The OP could state just how "open air" it needs to be?  I suspect you
>>could do a lot better if the user was expected to hold the phone up to
>>the speaker than if it had to work from across the room.
> 
> speaking of difficult channels
> 
> it would be really rewarding for someone to develop an emergency comm system
> for miners trapped underground...

There are portable systems for undeground communication. They are based 
on LF or VLF coupling to a wireline or rely on repeaters or some other 
underground infrastructure.


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

Eric Jacobsen <eric.jacobsen@ieee.org> wrote:

(snip, someone wrote)
>>it would be really rewarding for someone to develop an 
>> emergency comm system for miners trapped underground...

> There are standards for such things, and products that meet the
> standards.   I don't know how well any of them work in practice, but
> some seem to be pretty useful.

I used to know stories about systems for communicating with
submarines at either 15kHz or 7Hz.  The latter, as I remember it,
is the resonant frequency of the earth, such that one can't do
direction finding.   I don't know how deep you can get at
the lower frequencies with enough power.  Presumably the electrical
noise level is low down there (after a cave in with all power
turned off).

-- glen