On breadboarding Upper Sideband Syntonization at P=2 for space Radio Interferometry

Volodymyr Kudriashov

Serco Nederland B.V.

Testing concurring theories of gravity utilizing black holes' images requires two satellites maintaining coherence for hundreds of minutes at a sub-millimetre wavelength of 0.5mm. This is out of reach of space atomic clocks. We engineered a way out for this: the Upper Side Band Syntonisation concept, USBS. We have patented it. We have breadboarded it and tested with W-band radio interferometer instrument. We have demonstrated a coherence sufficient to operate at 0.5mm on timescale of 1 month. This is a game changer for space radio interferometry. Moreover, this is interesting for applications in communications and Earth Observation, Navigation. The USBS generates the sum frequency of two independent oscillators aboard two satellites of nominally the same frequency (but in practice slightly different) using a bi-directional link. This sum frequency under these conditions happens to be the same aboard both satellites, except for a rather minor Doppler residual difference, well tolerable for the application of black hole imaging. Compared against an ideally stable atomic reference clock external to the system, the sum frequency (realised with standard clocks, for example) would go its own way depending on the stability of the individual clocks, but, for our application of interest (and very many other applications) what matters is the relative coherence between the two receivers. We breadboarded USBS at P1 and P2 architectures, including the instrument-level demonstration. With a set up implementing the USBS with P1 at 103.2GHz we have maintained coherence up to ~1000s with Allan Deviation permitting interferometry towards 15THz, transport delays scale of 5 M km (with a potential to expand), transport asymmetry up to 60m. With P2 set-up at 46.3 GHz, we have demonstrated capability to operate at the nominally same RF tones in two counter-directions of inter-satellite link namely, the capability to carry-out radio interferometry at 0.5mm without a need to correct for the transport Doppler. We have maintained the coherence (correlation amplitude > 0.9) between two 92 GHz receivers driven by local oscillators (LOs) generated according to the USBS from two independent clocks A and B set at the same nominal frequency for 1 month without any sign of degradation (in fact the set up could have run for many more days with the same performance). By swapping a couple of cables to drive the two receivers with LOs generated by the two independent clocks A and B of the same nominal frequency but without using the USBS concept, the amplitude of the correlation vanished after just a few milliseconds, as expected. The plots of the phase show a peak-to-peak excursion mush less than in Harmony. The phase of the correlation, hence the relative phase between LOs (and receivers) in our RF lab USBS set up moves very little over a scale of a month, which gives us confidence in that phase calibration would be possible for a radio-telescope formed by a set of satellites implementing the USBS concept. We think that the USBS is robust against Doppler, i.e. even if satellites move relative to each other, the frequency syntonization (and hopefully the phase synchronisation) should be maintained as in the lab set up. Results obtained may be promising for the black hole imaging and other use cases.