Planar Array Diagnosis by Means of an Advanced Bayesian Compressive Processing

The detection of faulty radiators in planar antenna arrays is addressed in a probabilistic framework and it is solved through an innovative multitask Bayesian compressive sensing approach. More specifically, the proposed method formulates the failure detection problem at hand such that the real and imaginary parts of the sparse unknowns, which are the differential values of the excitations of the actual array with respect to those of the gold (i.e., failure-free) array distribution, are statistically correlated during the inversion of the measured data. Representative results from a wide set of numerical experiments are shown to assess the effectiveness of the proposed detection tool also in a comparative fashion with competitive state-of-the-art alternatives.