Can a beam of light be modulated close to it's frequency? I'm wondering if it would be possible to modulate a signal using photon pulses - where the frequency of the signal would be close to the frequency of the light. Also would it be possible to transmit multiple of these signals in parallel and have them not interfere with each other.
Thought experiment A:
I have a laser: green(500nm, 600kGhz) which can be controlled to emit individual photons. I feed a binary signal into this laser with the same frequency as the light frequency (600kGhz). For each 1 in the signal, the laser will emit a photon and for each 0 the laser will not emit a photon. Example: For an '1 1 0 1' bit stream - the laser is emitting photon A, waiting for the photon A to travel it's wavelength, it emits photon B, waits for photons A+B to travel their wavelength, does nothing, waits for photons A+B to travel another wavelength and then emits photon C. Using this scheme, I can transmit 600,000 GigaBits of information / sec to an observer 1 mile away by having a light sensor which would either sense a photon or not sense a photon at each wavelength interval.
Would this be possible ? How about if I emit/not emit a photon at 2*wavelength period ? What's the maximum signal frequency for which this would work ?
Thought experiment B:
Assuming the previous scenario works I have 2 signals + 2 lasers as described above :green(500nm, 600kGhz) & red(700nm, ~430kGhz). 
Can I pass the two photon 'beams' through an optical fiber (ignoring material properties of the fiber such as loss & lower light speed), have the mixed beams emerge 1 mile away, separate them using a prism and measure the original signal on each beam using a photon sensor ? 
Basically Would I be able to transmit a 600,000 GigaBit/s signal + a 430,000 GigaBit/s signal through an optical fiber using this method ? Or would the two signals somehow interfere with each other ?
Thanks!
 A: There are (at least) a couple problems with your proposals:
In experiment A, you start with "a laser ... which can be controlled to emit individual photons." However, such a laser is, as far as I know, impossible to make, because the stimulated emission process that generates the output in the laser is a random process, emitting photons at random times. This "shot noise" problem (although it's normally looked at as a receiver problem rather than a source problem) gives a well-known result in optical communications that it requires about 9 photons per bit to transmit a signal reliably.
As for experiment B, when you modulate a signal, the signal bandwidth is spread by the bandwidth of the modulation. So if you modulate a 600 THz signal with a 600 THz signal, the bandwidth of the modulated signal is now 600 THz. The initially "green" beam is now spread across the whole spectrum (from 300 - 900 THz). This effect will prevent demultiplexing your 600 THz signal (modulated at more than 50 THz) from your 650 THz signal at the receiving end.
