Ambiguity Resolution of Two Conformal Leaky-Wave Antennas via Deep Learning

Ambiguities in direction-of-arrival (DoA) estimation introduces significant challenges in wireless communication systems, particularly in applications requiring precise localization and sensing. These ambiguities can lead to misinterpretation of signal origins, severely impacting the performance and localization of systems such as Wireless Body Area Networks (WBANs) and Internet of Things (IoT) devices. In this paper, we investigate the ambiguities in an array of two symmetric conformal leaky-wave antenna (LWA) system, which, while offering enhanced field-of-view coverage, introduces complexities in resolving directional ambiguities. To address this issue, we propose a novel ambiguity resolution model based on a Long Short-Term Memory (LSTM) network. Our proposed LSTM-based approach achieves an accuracy of 99.99% in resolving ambiguities in comparison with other state of the art techniques. This advancement not only strengthens the reliability of conformal LWA systems but also lays the foundation for more robust and precise localization in wireless scenarios.