A metasurface antenna with beam steering capabilities for radio frequency energy harvesting

Limited control over parameters within the realm of design complexity, it is noted that there are incomplete unit cells displaying slightly lower transmittance in transmissive metasurface antennas. However, traditional methods require time-consuming iterative processes to select unit cells with multiple variables. Therefore, a novel approach is proposed for designing metasurface superstrates on patch antennas operating at 12.5 GHz, focusing on achieving beam steering capabilities and Radio Frequency (RF) energy harvesting functionality. It utilizes a Skeleton-based Attentions Neural Network (SANN) to aid in the complex metasurface unit cell design, enhancing the antenna's flexibility and potential applications. This network is trained to generate patterns for the metasurface unit cell, facilitating accurate and efficient design. The key feature of this antenna is beam steering in multiple directions. Steering is achieved by rotating the metasurface plane in an anti-clockwise direction. Experimental results show that by rotating the metasurface plane at different angles, the beam is directed towards specific directions, such as 10 degrees, 15 degrees, 18 degrees, 21 degrees, and 23 degrees. A load resistance of 8k ohm and a full wave rectifier with a Positive-Intrinsic-Negative (PIN) diode are utilized to obtain a Direct Current (DC) output voltage of 1.1 V in the proposed rectifier. Copper material is used for the metasurface plane, and the antenna is measured using RT duriod.