Mixing digital signals of same frequency to a frequency the sum of the input

[AFBlumen]

Hello Guys. Good day
I wanted to mix 4 1GHz digital signals to a single 4GHz signal.
Let's name them A, B, C, D
so every a nanosecond they pulse one time. And I plan to multiply the four input signals by 4.
So they will pulse 4 times per nanosecond and using a clock to time them.
So if the input in input A is A1, A2 the output is A1, A1, A1, A1, A2, A2, A2, A2.
Then,
A1, A1, A1, A1
B1, B1, B1, B1
C1, C1, C1, C1
D1, D1, D1, D1
with the clock will be A1, B1, C1, D1 in 4 times the input frequency.
Am I right there?
Thank you!

[MagicSmoker]

If the 1GHz pulses have a 50% duty cycle and are phase-aligned then, no, you can't combine them. If you can reduce the duty cycle to much less than 25% (e.g. - 10%) and phase-shift them by 90 deg. from each other (e.g. - 0, 90, 180, 270) THEN you could combine them with OR gates (perhaps even with Schottky diodes, though I'm more of a power electronics guy than digital, much less 1GHz+ stuff).

Draw this out on a piece of paper and you'll likely see why.

[David Hess]

Some low noise frequency doublers work exactly like that.  A quadrature signals ( 0 and 90 degrees or -45 and +45) is produced from a single input and fed to a mixer which adds them together to double the original input. 

If the single frequency inputs to the mixer are not held in quadrature, then a DC offset at the output is produced.  So assuming the 4 inputs are phase coherent, filter the DC output from the mixer, compare it to zero, and use that to drive a voltage variable phase shifter at one input.

If the sources are *not* phase coherent, like from separate oscillators, then do the same thing but control the phase and frequency of one oscillator.