UV-range

Two important parameters to determine for deriving antenna based complex gains in each cycle will be the values of $\Delta B$ and the antennas to be used. Including all baselines which fall within the selected uv-range in the uv-plane will certainly not be wise. This may result into an under-determined set of equations (i.e, may include antennas, the complex gains for which cannot be solved for). Rigorously speaking, visibilities from a set of antennas, each one of which satisfies at least one amplitude closure relation, constitute a solvable set of equations. This criteria can be used to determine the set of antennas to be used for a given uv-range.

$\Delta B$ can be determined by demanding that the uv-range at each iteration be increased by an amount which includes $N$ more antennas than used in the previous iteration, all of which satisfy $C$ amplitude closure relations. The values of $N$ and $C$ can be imperially determined. $C$ will be function of the signal-to-noise (SNR) ratio in the data. The antenna based calibration process essentially determines a consistent set of antenna gains by averaging all the available closure relations. More closure relations reduces the error budget on the solutions. For high SNR data (which will be case for most of our observations), the process is expected to be weakly dependent on $C$. $N$ can be determined by some heuristic argument (for moderately resolved sources, the convergence of the process will only weakly depend on this too).