How exactly is digital information sent with radio waves?

[Edi]

What I am interested in here specifically is - how are bits represented?
Is it as "1" is represented with a sent photon [at those specific frequencies] and "0" with a .. silence (no photon sent) ?
... but this, in that case, would get really complicated, because other photons from other sources would interfere and the "silence" would easily be filled with something.
.. or is this avoided by making the silence not necessarily as long [think wavelength] as the photons, so that, even if disturbance occurs, it cannot possibly be interpreted to be from the same signal as the timing would not mach they would overlay .. so only the photon that comes i at specific time is registered..
im just really just thinking out loud here as I write.
But this bothers me. At a fundamental level.

 

[berkeman]

What I am interested in here specifically is - how are bits represented?
Is it as "1" is represented with a sent photon [at those specific frequencies] and "0" with a .. silence (no photon sent) ?
... but this, in that case, would get really complicated, because other photons from other sources would interfere and the "silence" would easily be filled with something.
.. or is this avoided by making the silence not necessarily as long [think wavelength] as the photons, so that, even if disturbance occurs, it cannot possibly be interpreted to be from the same signal as the timing would not mach they would overlay .. so only the photon that comes i at specific time is registered..
im just really just thinking out loud here as I write.
But this bothers me. At a fundamental level.

The concept you are asking about is "modulation":

http://en.wikipedia.org/wiki/Modulation

.

 

[fleem]

There is a whole science involving how one best stores or transmits information through a noisy channel. But the simple answer to your question is this: a reasonable modulation method is chosen (might be the on-off method you described, or might be some other modulation method) and then one must simply make sure there is enough power (loudness) in the signal to make it possible for the receiver to hear it among the noise. In practice (for a typical radio link) there will be guzillions of photons for each bit of information. For example, if your modulation method were used, the transmitter sends a guzillion photons in a split second to mean a "1" bit, but no photons to mean a 0 bit. This is done because a single photon isn't nearly loud enough for the receiver to hear it above the noise.

 

[Edi]

mm, ok, that explains quite much. [that one photons is to .. weak] I made this fast calculation and, with a frequency of about 1000MHz, we could send more than 100MB per second.. that is quite fast. And could potentially get up to 200MB (mega frikin bytes, not bits) if one just makes the "silences" short enough.
More, if the frequency is higher..

so, one photon is not enough, but.. I heard of these "super photons" when a bunch of them travel as a single thing.. If we could learn how to generate these things efficiently, we could maybe get up to those speeds mentioned above?
That would be insane [..ly cool]

 

[fleem]

A single electrical pulse of just a few watts sent into an antenna does produce guzzillions of photons.

You can actually send more than one bit per single wave of the signal. For example, you might define your modulation where there might be 32 different values for the amplitude of a single wave of a signal, where each amplitude value represents a different pattern of 5 bits (2^5=32). So ten waves could send fifty bits. See how the frequency of the carrier signal is not a limit on the data rate?

There is a theoretical maximum data rate that can be sent through any channel, and it depends on the signal power (your transmit power), the frequency bandwidth occupied by the signal (You can send more data through a 5MHz channel than you can through a 1MHz channel even with the same signal power), and how much noise is in the channel. In fact, if there is no noise in the channel then you can send with an infinite data rate (assuming you use a clever modulation). for example you could have thousands of amplitude levels for each wave of the signal where each one means a different sequence of bits (you could only do that if there were very little noise). You could do the same thing with the time that the wave is sent (phase modulation). That's why a 56K modem can work in a 3KHz bandwidth phone line. phone lines are typically fairly low noise.