Application of the Compressive-Sensing Approach to the Design of Sparse Arrays for SATCOM Applications

Current SATCOM systems employ multiple reflectors with a one-feed-per-beam configuration to synthesize narrow spot-beams. However, these systems are very complex and offer very limited reconfigurability. Active antenna arrays are attractive solutions [1], although are often expensive due to the large number of elements and electronic components involved. Aperiodic array antennas can substantially reduce the number of elements and costs with respect to regular arrays but their design is challenging [2]. Several synthesis methods have been proposed, yet aperiodic array design techniques are not as mature as those in use for their regular array counterparts. These methods are often either: (i) accurate but computationally expensive (e.g. Genetic Algorithms [3]), or; (ii) efficient but simplified (e.g. Density Tapered method [4]). Compressive Sensing (CS) has been recently applied to the synthesis of sparse antenna arrays. The method can optimize large maximally sparse antenna array problems in a fast, deterministic and flexible way [5]. In previous research publications, the authors have (i) extended the original formulation to the multi-beam scenario; (ii) exploited array layout symmetry and modular design; and (iii) hybridized the original iterative optimization procedure with a full-wave EM analysis, so as to include the effects of mutual coupling into the design process and studied for arrays of strongly coupled antennas elements, such as dipoles, as well as large planar arrays of pipe horns [6, 7]. Additionally the authors have addressed multi element type design [8] and, more recently, are investigating reconfigurable arrays (i.e. arrays designed to provide a set of arbitrary-shaped beams) and isophoric arrays (i.e. arrays with a single excitation amplitude). The main directions are summarized in Fig. 1