Sparse Array Synthesis Including Mutual Coupling for MU-MIMO Average Capacity Maximization

Stochastic array antenna figures‐of‐merit for quality‐of‐service‐enhanced massive MIMO

“…However, when using () to design array antennas, applying these conventional normalization techniques partially hides essential aspects like angle‐of‐arrival (AOA)‐dependence, MC, and impedance matching. This follows from the fact that, when designing physical array antennas based on a QoS‐metric like the SINR, the impact of the antennas is typically embedded in the channel vectors $$\mathbf {h}_1,\dots,\mathbf {h}_K$$ by means of, for example, embedded element patterns (EEPs) or a mutual coupling matrix (MCM), see, for example [2, 5]. In other words, conventional power‐based normalization hides gain fluctuations induced by a (possibly poor) array design, as they are implicitly compensated for by the UE transmit powers.…”

“…Recently, however, communication-oriented array design [1] has shown that unconventional layouts can enhance the user equipment (UE) QoS. Examples of considered QoS-based FoMs are ergodic channel capacity [2], SINR [3,4], or SINR-dependent metrics like bit error rate [5] and ergodic sum rate [5][6][7][8]. However, understanding the physical phenomena behind array-layout-induced QoS improvements is complicated, as is illustrated by, for instance, conflicting statements on whether mutual coupling (MC) enhances or deteriorates ergodic channel capacity; see, for example [9] and references therein.…”

“…. , h K by means of, for example, embedded element patterns (EEPs) or a mutual coupling matrix (MCM), see, for example [2,5]. In other words, conventional power-based normalization hides gain fluctuations induced by a (possibly poor) array design, as they are implicitly compensated for by the UE transmit powers.…”

Controllable Local Propagation Environment to Maximize the Multiplexing Capability of Massive MIMO Systems

“…Sophisticated beamforming and precoding were studied in [18][19][20]. The use of irregular arrays for MaMIMO systems was discussed in [21,22], and several synthesis algorithms have been proposed in the literature to solve this task effectively [23][24][25][26][27][28]. Another promising approach is the abandoning of the "cell" concept in the so-called cell-free massive MIMO [29][30][31].…”