Application of Krylov based methods in calibration for radio astronomy

As the number of antennas in the modern radio-telescopes increases, the computational complexity of the calibration algorithms becomes more and more important. In this paper we use the Khatri-Rao structure of the covariance data model used for such calibrations and combine it with Krylov subspace based methods to achieve accurate calibration results with low complexity, very small memory usage and fast convergence properties. We also demonstrate the proposed method on experimental data measured by the LOFAR radio-telescope.