Computer Vision Enabled Calibration of Additively Manufactured Conformal Phased Arrays Utilizing 3-D Depth Sensing Camera

The development of flexible phased array systems has drawn significant interest due to their adaptive beam steerability and deployable structures to support various platforms virtually on every conformal surface. Additively manufactured tile-based phased array offers a lightweight, flexible, and massively scalable solution with reduced cost and fabrication time. The main challenge for such conformal phased array systems is to maintain array performance under various deformations, which requires calibration of the amplitude and/or phase distribution of the antenna elements. To address this challenge, a computer vision-enabled on-the-fly adaptive shape calibration and phase correction method is proposed. The authors introduce the usage of smartphones, with integrated 3-D depth cameras and infrared (IR) sensors, and a novel computer vision algorithm to detect the bend angles between neighboring subarray tiles. A <inline-formula> <tex-math notation="LaTeX">$2 \times $ </tex-math></inline-formula> additively manufactured tile-based flexible phased array is utilized as a proof-of-concept (POC) prototype to demonstrate the calibration approach. The proposed algorithm achieves a very good <1° angular prediction accuracy and demonstrates successful calibration of the phased array under 15°, 30°, and 45° bend angles under both symmetrical and asymmetrical configurations with improvements of gain as much as 7 dB. The calibrated bent phased arrays can also achieve a maximum steering range of 110°. This approach presents a highly accurate and cost-effective calibration process that can enable massive fabrication and implementation of tile-based flexible phased arrays for next-generation 5G/mmWave wearable and conformal smart skin, Internet of Things (IoT), Industry 4.0, and massive multiple-input multiple-output (MIMO) applications.