Reconfigurable adaptive array beamforming by antenna selection

Traditional adaptive array beamforming with a fixed array configuration can lead to significant inefficiencies and performance loss under different scenarios. As antennas become smaller and cheaper relative to front-ends, it becomes important to devise a reconfigurable adaptive antenna array (RAAA) strategy to yield high signal to noise and interference ratio using fewer antennas. This is achieved by selecting $K$ from $N$ antennas to minimize the Spatial Correlation Coefficient (SCC) between the desired signal and the interference. The lower bound of optimum SCC is formulated with two relaxation methods to give information about the suitable number of selected antennas K. A Correlation Measurement (CM) method is proposed to select the optimum subarray with K antennas, thereby reducing complexity. We carry out performance analysis and show that a 1/K²-suboptimum solution can be guaranteed with arbitrary shaped arrays. Furthermore, a Difference of Convex Sets (DCS) method is proposed to select the optimum subarray with controlled quiescent pattern in order to reduce the effect of interference during the reconfiguration time. The utility of the proposed array reconfiguration for performance improvement without increasing the cost is demonstrated using both simulated and experimental data.