Near-Field Phased Array Diagnostics by a Subspace Projection Method

This paper aims to introduce a new strategy for the detection of faulty elements in phased array antennas. The approach takes advantage on the possibility in phased array to introduce the steering diversity. Accordingly, by collecting the radiated field over the measurement aperture while steering the beam, an approach inspired by the Time reversal MUSIC (TR-MUSIC) algorithm is exploited to achieve the diagnostics. At first, it is shown that such an algorithm allows the detection of only completely turned-off elements as long as the number of defective elements is relatively (depending on the configuration parameters) high. To overcome such a limitation, the difference model is built and the approach succeeds in detecting any failure (amplitude and phase failure) by a reduced set of data, mainly depending on the number of faulty elements. In addition, from the mere knowledge of the geometric parameters of the measurement configuration and by invoking the theory of degrees of freedom, the conditions for which the algorithm succeeded in the diagnostics are analytically derived. Furthermore, it is highlighted that the proposed algorithm achieves the diagnostics even when both the spacing between the array elements and the element factor is not a priori known. An extensive Montecarlo numerical analysis is included to show the capability of the approach to achieve a high probability of detection with a reduced set of data. Finally, full-wave simulations are also included.