Unmanned Aerial Vehicles (UAVs) are rapidly used in a variety of applications such as surveillance, search and rescue operation, delivery, etc. These operations strongly depend on the performance of the air-to-ground (A2G) communication link. This link communicates vital information between the UAV and the ground station (GS). However, due to the UAV’s unpredictable mobility, the A2G link becomes unreliable, which limits the quality of data transfer. An Adaptable Radar Antenna Array (ARAA) can be used to solve this problem by effectively handling transmitted data through altering the antenna pattern using the beamforming process. This work proposes a new algorithm for selecting antenna patterns with proper beamforming by considering the effect of wobbling and mutual coupling (MC). In addition, the proposed algorithm is used to reduce the maximum sidelobe level (SLL) and optimize the distance between the interelement of ARAA, which directly controls the MC effect in a time-varying channel. Also, an analysis is done for various antenna configurations, such as Uniform Linear Array (ULA), Uniform Rectangular Array (URA), and Uniform Circular Array (UCA), while varying different UAV parameters like altitude, speed, location error, and beamforming capacity.