Efficient Beamforming with Multi-Active Multi-Passive Antenna Arrays

In this paper we consider Multi-Active Multi-Passive (MAMP) antenna arrays with a reduced number of active elements (AEs) for emulating the patterns of all-active uniform linear arrays (ULAs), which leads to a highly nonlinear task. Moreover, we present a novel alternating optimization stochastic beamforming algorithm (AO-SBA) for the joint calculation of the passive elements' (PEs) loads and baseband weights of the MAMP arrays. We consider both linear and rectangular array configurations, which are modeled by taking into account all coupling effects. We extend state-of-the-art methods that were originally designed for single-fed load controlled antennas and compare them against our proposed iterative scheme through an extensive set of experiments. Our results indicate that MAMP arrays produce a radiation pattern similar to that of a ULA with accuracy up to 98.9%, while the number of AEs is reduced by 50% and in some cases with a significant reduction of the side lobes. Thus, the complexity and cost of the antenna array can be reduced without compromising the quality of the resulting beam. Given the limited prior state-of-the-art in this field, we believe that this work will pave the way for MAMP arrays to become a strong contender for future multi-antenna systems that require high spatial selectivity with specific hardware, size and cost constraints.