New routine for Resonator Identification

[aorgazf]

Hi, The characterisation of a QPU from scratch, requires sweeping large portions of the spectrum to find resonators. The resonator spectroscopy routine that we currently have is not well suited for that purpose. It works well to extract the resonator frequency when the frequency is already know and the scan is narrow around the resonance frequency. But for wider scans like this, when multiple resonators are found, it fails:

Another problem is that the width of the spectrum to be swept is limited to the bandwidth of the control instruments. Ideally we could have a function that when provided a wider section of the spectrum it would perform multiple sweeps in smaller sections of it. If the platform uses local oscillators, their frequency will need to be shifted for every section.

The current fitting oftentimes fails due to the presence of standing waves. Those standing waves do not affect the phase information, though:

Using the phase to detect the resonators should be less prone to invalid results.

There is an alternative way to find and identify resonators:

• perform a wide resonator spectroscopy (splitting it into smaller sweeps)
• unwrap phase
• derive the phase
• find maxima (each maximum corresponds to a resonator)
• once all resonators are found, apply flux to each qubit flux line separately and see which of those resonances moved (that allows the identification of the resonance with the corresponding qubit.