Local array calibration using parametric modeling of position errors and a sparse calibration grid

We propose a new method for offline calibration of a sensor array with respect to position, gain and phase errors of the sensor elements. We model the effect of position errors explicitly, such that they can be corrected parametrically using a least-squares formulation. This is equivalent to the already existing Maximum-Likelihood position estimator, but does not require a minimal number of calibration sources. While this is crucial, especially for large (e.g. 2D) arrays, it also reduces costs by using a sparse calibration grid for smaller arrays. The effectiveness of the method is demonstrated by simulations and based on a real data experiment.