In-situ mixer sideband calibration with superconducting qubits

Precise microwave control is critical for the optimal performance of superconducting quantum processors. We present an in-situ sideband calibration technique of IQ mixers with superconducting qubits, enabling the diagnosis and elimination of unwanted image-sideband signals without additional hardware. Although such calibration could be achieved by adjusting the qubit frequency to be resonant with the image sideband and minimizing the Rabi oscillation driven by the image sideband, it is inherently limited by the qubit's flux-tuning frequency range. To overcome this limitation, we propose a novel strategy to learn and correct imperfect mixer phases. Our experimental validation confirms the effectiveness of this protocol in precisely canceling the image sideband, enhancing quantum processor functionality.