We propose a new approach to the mask-constrained synthesis of shaped beams through generic array antennas having arbitrary layouts and element patterns. The introduced general methodology allows casting the design problem as a finite number of Convex Programming optimizations and can generate different excitation solutions all fulfilling the required specifications. Numerical examples, including planar arrays and the full-wave synthesis of an aperiodic array, are provided in support of the given theory.

We present a new approach to the problem of detecting element failures in any kind of array antenna. The proposed method requires only phaseless information on the generated field and reduces as much as possible the number of needed measurements by exploiting the Compressed Sensing (CS) theory. The solution procedure works regardless of whether the probes are located in the near-field or far-field region of the antenna under test, and it is the first and only CS-based technique able to perform the diagnostics by exploiting only phaseless measurements taken exclusively in the near-field zone. It also results simple and fast as it requires a single measurement surface and reduces, in most practical instances, to the solution of a convex programming problem. The given theory is supported by very many numerical experiments involving full-wave electromagnetic simulations of realistic arrays and considering both scalar and vector fields with different types of scanning surfaces and different noise levels. Both ON-OFF and phase faults are dealt with. An experimental validation on a fabricated array is also presented.

With reference to the mask-constrained power synthesis of shaped beams through fixedgeometry antenna arrays, we elaborate a recently proposed approach and introduce an innovative effective technique. In particular, the proposed formulation, which can take into account mutual coupling and mounting platform effects, relies on a nested optimization where the external global optimization acts on the field's phase shifts over a minimal number of 'control points' located into the target region whereas the internal optimization acts instead on excitations. As the internal optimization of the ripple is shown to result in a Convex Programming problem and the external optimization deals with a reduced number of unknowns, a full control of the shaped beam's ripple and sidelobe level is achieved even in the case of arrays having a large size and aimed at generating large-footprint patterns. Examples involving comparisons with benchmark approaches as well as full-wave simulated realistic antennas are provided.

In hyperthermia treatments, the possibility of generating a specific absorption rate distribution peaked into the tumor and bounded elsewhere represents the main clinical need. As multi-frequency applicators can be exploited to reduce the occurrence of undesired hot-spots, an optimal multi-frequency approach is proposed and discussed. Being formulated in terms of a convex programming problem, the globally optimal solution can be determined (only for very special cases). The procedure, presented for the case of scalar fields, can be extended both to the case of vector fields and to the more interesting problem of shaping (rather than just focusing) the specific absorption rate distribution.

We introduce a novel approach for the mask-constrained power synthesis of 1-D fields enabling near-field shaping according to arbitrary masks. This method presents an innovative framework based on the spectral factorization technique and a warping strategy, allowing for an accurate representation of the near field as a minimally redundant bandlimited function. This provides valuable theoretical tools to effectively solve the problem. In fact, the synthesis is formulated as a linear programming problem followed by a polynomial factorization. Additionally, the approach allows for the a-priori identification of the minimum size of the source required to fulfill a given near-field intensity mask.

The phase retrieval problem is dealt with for the challenging case where just a single set of (phaseless) field data is available. In particular, the adoption of a new set of intersecting curves in the spectral plane allows decisive improvements over our recent approaches emulating the solution of crosswords puzzles, thus removing relevant limitations on the source size. Numerical examples support the given theory and confirm the effectiveness of the developed procedure.