Title:Learning to Calibrate Quantum Control Pulses by Iterative Deconvolution

Abstract:In experimental control of quantum systems, the precision is often hindered by the imperfection of applied electronics that distort control pulses delivered to the target quantum devices. To mitigate such error, the deconvolution is commonly used for compensating the distortion via an identified convolutional model, and its effectiveness is limited by the model imprecision (e.g., parameter error or unmodeled nonlinear distortion). In this paper, we propose an iterative scheme that calibrates the pulse by repeatedly applying the deconvolution operation to the error signal. Numerical simulations show that the resulting iterative deconvolution method can correct both linear and nonlinear model errors to an arbitrary precision at the sampling times. Moreover, we find that residue errors between the sampling points are invisible and hence not correctible by iterative deconvolution, but it can be suppressed by actively introducing nonlinear components in the control transmission line.