I. Introduction
Sensor arrays are a key technology with several applications in radar, sonar, microwave imaging, medical ultrasound, and wireless communications, to list a few [1]. The many advantages of arrays include high signal-to-noise ratio (SNR) gain, spatial diversity, and the capability to cancel interferences by beamforming. The ability to resolve targets improves with increasing aperture, which encourages using short carrier wavelengths. This allows for designing electrically large arrays with small form factors by packing a very large number (on the order of hundreds) of elements into a tiny (on the order of a cm) physical area. On the other hand, the cost, power consumption, and computational load commonly associated with signal processing for many antenna elements and dedicated transceiver chains may become prohibitively large. These issues are especially pronounced for fully digital arrays, where each array element is connected to a separate front end, which includes radio and intermediate frequency (RF-IF) components and an analog-to-digital converter (ADC) or a digital-to-analog converter (DAC). For example, a planar antenna array operating in the THz frequencies of the radio spectrum may in principle even fit thousands of elements in an area of only a few square centimeters. The practical applicability of such fully digital systems is limited by the number of required RF-IF front ends, and the typical high sampling rates and bandwidths imposed on the DACs/ADCs.