# Interleave two analog signals

I am trying to interleave two similar but not identical pass-band (analog) signals (i.e., both have a center frequency of 15 MHz, and a bandwidth of 200 kHz) on a sample by sample basis. The idea is that once interleaved in the analog domain, my one-channel receiver will down-convert the interleaved signal to baseband, and will sample it at the desired rate. Then, I will be able to de-interleave the signal on a sample by sample basis in the digital domain to recover the two signals. Pictorially, what I am trying to achieve in the analog domain is depicted below:

Analog Domain

S1 (pass-band signal): + + + + + +
S2 (pass-band signal): o o o o o o
Interleaved (pass-band signal, to Receive Channel) : + o + o + o + o + o


The digital part is easy to realize, however, I have few issues with the analog part. What is the best way to interleave two similar pass-band (analog) signals such that when they are down-converted to baseband and sampled, they can be de-interleaved on a sample by sample basis as shown above?

I thought about one way to achieve this through using an RF switch with a switching speed twice the center frequency. Does that make sense? Are there better ways to realize this?

I'll appreciate any ideas.

• I am fairly sure that knowing how you modulate your carrier is fundamental. – Vladimir Cravero Apr 2 '15 at 19:46
• have a look at "time division multiplexing" anyway, and be warned: your resulting signal power spectrum will probably be wider than 200Hz and possibly carry some power at 30MHz or other frequencies. – Vladimir Cravero Apr 2 '15 at 19:51
• How do you plan to synchronize the analog interleaver with the digital sampler? Are you at all familiar with how broadcast FM stereo works? – Dave Tweed Apr 2 '15 at 20:27
• And what is "the desired" sampling rate? – Brian Drummond Apr 2 '15 at 20:51
• You need to pop up a couple of levels and explain what you are really trying to accomplish. Your imagined solution has some serious problems. It would be easier to describe how to attack the overall problem than to have to start by debunking what you are trying to do. For one thing, you don't seem to be aware of the bandwidth implications of multiplexing as you describe, let alone how to recover the clock as Dave already pointed out. – Olin Lathrop Apr 2 '15 at 23:23