Researchers at Digital University Kerala have developed an artificial intelligence (AI)-driven system that can identify antenna types and detect performance faults without dismantling or physically testing them.
The breakthrough holds potential in transforming communication, defence and aerospace industries. The innovation could also lead to self-diagnosing communication modules in satellites, drones and Internet of Things (IoT) devices to enable real-time performance monitoring and fault detection.
The research, published in the IEEE Journal of Microwaves, was led by Anitha Gopi, Sruthi Pallathuvalappil, Elizabeth George and Alex James. The study introduces a neuro-memristive 3D crossbar system that is capable of reading antenna radiation patterns like images. It also classifies the antenna type (dipole, monopole or path) using an advanced 3D memristive convolutional neural network (3D-CNN).
Traditionally, antenna testing involves costly and time-consuming experiments inside anechoic chambers (rooms designed to stop reflections or echoes of either sound or electromagnetic waves).
According to the researchers, the new method involves pixel sampling and AI algorithms to analyse electromagnetic field data, thereby cutting down on power, area and testing time. Hardware implementation was done using the Skywater 130-nanometre process, an open-source semiconductor platform.
“Our system offers a compact, non-invasive way to ensure antennas are working correctly, even in noisy or harsh environments,” Prof. James, the corresponding author, said. This is especially valuable for defence and remote communication applications, he added.
The study also compared the 3D-CNN with other machine learning models such as YOLOv8 and VGG-19 to find that the neuro-memristive approach delivered higher accuracy and faster processing even under signal noise like Gaussian and white noise.
Ms. Gopi said the research bridges AI and hardware innovation to make antenna testing smarter, faster, and more reliable.
Published – October 19, 2025 07:46 pm IST
