I would suggest that any signal would be frequency modulated to remove the Doppler component of the transmitting station's planetary motion, of both rotation and revolution. Of course, it is possible the transmitter might not be on a planetary surface at all, but in orbit around the star system where the ETs are set up.

I am assuming the signal would be a tight beam aimed at a particular target star, in which case the frequency shift caused by the relative motion of both the transmitter and receiver would be taken into account, and removed from the signal. Of course, if ET is broadcasting omnidirectionally, or into a broad area of the sky, he would have no way of knowing how to Doppler-correct the signal for motion.

But I still think that using a harmonic or an transcendental number multiple of the frequency would satisfy both criteria: that the signal be obviously artificial, but still related to natural cosmic frequency. Of course, any motion, whether around a planet's axis, a planetary orbit, or even the orbit of the star around the galaxy, would Doppler- shift the signal and require a broad-band strategy at the receiver.

Another useful concept is the "waterhole", the space between the natural emission lines of the H+ and OH- radicals, also a common signal in the molecular clouds in the galactic plane. Both these emission lines straddle the optimum low-noise saddle in the radio background, so they would be a good place to listen. Besides, all living things gather around the waterhole, don't they?

Yeah, its a needle in a haystack problem, all right.