Calibration of the SKA-low antenna array using drones

Due to the always higher demand for precision and survey speed coming from the astronomers, it is required to lean on always bigger radio telescopes. The latter need accurate calibration to ensure the capture of high quality images. In particular, the Square Kilometre Array (SKA), the world's largest radio telescope currently in development with expected deployment by late 2020s, will need precise and accurate calibration. Natural calibration, e.g. Self-Cal algorithm, is widely used in the radio astronomy literature and allows the calibration of several SKA pathfinders, e.g. the Very Large Array (VLA) and Murchison Widefield Array (MWA). The increased sensitivity and the very large number of antennas of the SKA reduce the hope of a sole self-calibration. Research with artificial sources carried by a Unmanned Aerial Vehicle (UAV) has already led to successful far-field calibrations of radio telescopes. In this work, we show a near-field calibration procedure applied on a SKA station composed by 256 log-periodic antennas. A comparison is also made with a far-field calibration procedure. Two antennas mounted on a UAV act as calibration sources. Several flight strategies are studied with uncertainties on the position and UAV attitude. A maximum error between exact and modeled pattern of 40 dB with respect to the main lobe is obtained for a Gaussian noise of standard deviation of 5 cm. We show that near-field calibration enables to correctly model embedded element and array pattern. In comparison with a far-field proceeding, the flight distance is substantially reduced and the collected data carry more information. Near-field calibration offers a new window which has not been yet fully explored and is expected to give promising results.