The emerging reconfigurable antenna (RA) array technology promises capacity enhancement through dynamic antenna positioning. Traditional approaches enforce half-wavelength or greater spacing among RA elements to avoid mutual coupling, limiting the solution space. Additionally, achieving sufficient spatial channel sampling requires numerous discrete RA positions (ports), while high-frequency scenarios with hybrid processing demand many physical RAs to maintain array gains. This leads to exponential growth in the solution space. In this work, we propose two techniques to address the former challenge: (1) surrounding a limited number of active RAs with passive ones terminated to tunable analog loads to \textit{exploit} mutual coupling and increase array gain, and (2) employing tunable loads on each RA in an all-active design to \textit{eliminate} mutual coupling in the analog domain. Both methods enable arbitrary RA spacing, unlocking the full solution space. Regarding the latter challenge, we develop greedy and meta-heuristic port selection algorithms, alongside low-complexity heuristic variants, that efficiently handle over $10^{20}$ array configurations. Furthermore, we optimize the loading values to maximize the sum-rate in a multiple-input single-output broadcast channel under transmission power constraints, assuming a heuristic linear precoder. In addition, we analyze performance degradation from quantized loads and propose corresponding robust designs. Numerical simulations reveal 20-56\% sum-rate gains over benchmarks and around 60\% performance recovery under quantization errors.

翻译：新兴的可重构天线阵列技术有望通过动态天线定位来提升信道容量。传统方法要求可重构天线单元间距至少为半波长以避免互耦效应，从而限制了解空间。此外，实现充分的空间信道采样需要大量离散的可重构天线位置（端口），而采用混合处理的毫米波场景则需要大量物理天线来维持阵列增益，这导致解空间呈指数级增长。针对前者挑战，本文提出两种技术：（1）在有限数量的有源可重构天线周围布置端接可调模拟负载的无源天线，利用互耦效应提升阵列增益；（2）在全有源天线设计中为每个可重构天线配备可调负载，在模拟域消除互耦效应。两种方法均能实现任意天线间距，从而释放完整解空间。针对后者挑战，我们开发了贪婪算法和元启发式端口选择算法及其低复杂度启发式变体，可高效处理超过10^20种阵列配置。此外，在假设启发式线性预编码器条件下，我们通过优化负载值来最大化发射功率约束下多输入单输出广播信道的和速率，同时分析了量化负载带来的性能损失并提出了相应的鲁棒设计方案。数值仿真显示，相较于基准方案实现了20-56%的和速率增益，且在量化误差下仍有约60%的性能恢复。