[mdlug] Curious - Phone Tapping Tech

[Michael Mol]

On Thu, Jun 21, 2012 at 1:49 PM, Ingles, Raymond
<Raymond.Ingles at compuware.com> wrote:
>> From: Michael Mol
>> On Thu, Jun 21, 2012 at 1:32 PM, Aaron Kulkis <akulkis00 at gmail.com> wrote:
>
>> > Every current travelling "outbound" on one wire of a twisted pair
>> > has an equal and opposite "inbound" current on the other half of the pair,
>> > making the net current going through the loop = Zero, and therefore, no
>> > way to detect ANYTHING  within the cable bundle.
>>
>> That assumes the two halves of the pair are coaxial, which they're not.
>
> Right. The further away from the twisted pair, the closer the signal resembles the coaxial case, but there *is* some leakage. In practice, though, differentiating between multiple twisted-pair signals would be quite the chore. I suppose if you had some idea of the signaling protocols involved to establish some heuristics, and an FPGA or similar dedicated hardware to devote to DSP, you *might* be able to extract a signal, but it'd likely be rather noisy.

The case in question is just AM-encoded voice without a carrier
signal...voice comm over POTS. The human ear should be able to cope,
especially once you use some of the filtering I described above.

But, again, it'd likely be easier to just make a phone call on the
adjacent pair, and let crosstalk handle things for you.

-- 
:wq