Array gain/phase calibration techniques for adaptive beamforming and direction finding

A calibration technique is proposed for handling adaptive beamforming and bearing estimation problems involving unknown perturbed sensor gain and phase. This calibration technique is applied on the MUSIC estimator, assuming that two or more signal sources (in which the direction of arrival of one or two of the signal sources are known temporarily) exist in the signal field, so as to estimate the true sensor gain and phase. The basic idea of the technique is to apply the first-order Taylor series expansion to approximate the true array steering vector from the nominal one. A set of linear equations is then formed, using the null characteristic of the MUSIC spectrum, from which the error steering vector (the difference between the actual steering vector and the nominal steering vector), which contains the gain/phase information of the array sensors, can be solved for. This technique exhibits relatively stable performance compared with existing techniques in the sense that it produces the required estimates consistently without the need for iterative computation and initialisation. This is illustrated with numerical results obtained from several Monte Carlo experiments.<>