NEAR: Neural Electromagnetic Array Response

Yinyan Bu, Jiajie Yu, Kai Zheng, Xinyu Zhang, Piya Pal

Abstract: We address the challenge of achieving angular super-resolution in multi-antenna radar systems that are widely used for localization, navigation, and automotive perception. A multi-antenna radar achieves very high resolution by computationally creating a large virtual sensing system using very few physical antennas. However, practical constraints imposed by hardware, noise, and a limited number of antennas can impede its performance. Conventional supervised learning models that rely on extensive pre-training with large datasets, often exhibit poor generalization in unseen environments. To overcome these limitations, we propose NEAR, an untrained implicit neural representation (INR) framework that predicts radar responses at unseen locations from sparse measurements, by leveraging latent harmonic structures inherent in radar wave propagation. We establish new theoretical results linking antenna array response to expressive power of INR architectures, and develop a novel physics-informed and latent geometry-aware regularizer. Our approach integrates classical signal representation with modern implicit neural learning, enabling high-resolution radar sensing that is both interpretable and generalizable. Extensive simulations and real-world experiments using radar platforms demonstrate NEAR's effectiveness and its ability to adapt to unseen environments.

Lay Summary:

MIMO antenna systems have emerged as a crucial sensing modality for advanced sensing tasks such as driver assistance systems (ADAS) and autonomous vehicles, especially due to its robustness to adverse weather conditions. However, practical constraints imposed by hardware, noise, limited budget of antennas, and dynamic environment can impede its sensing performance. Can we achieve super-resolution sensing using single-shot data under sample-starved conditions and stringent hardware constraints?

Instead of building new hardware, we developed new algorithms. Inspired by implicit neural representations (INRs), we propose NEAR, a method that models each coordinate in the antenna space as a continuous function representing the expected EM wave response at that point. Guided by the physics of planar wave propagation and the underlying harmonic structure, NEAR learns these functions directly from single-shot measurements collected by sparse antennas, and predicts the EM response at arbitrary, unseen locations.

Overall, we believe our findings contribute to advancing research in INRs and their unique applications in super-resolution active sensing. Our work also marks the first step towards leveraging INRs for predicting unseen antenna responses in MIMO antenna sensing, paving the way for new opportunities to enhance the performance of future sensing and localization systems.

Application-Driven Machine Learning: This submission is on Application-Driven Machine Learning.

Primary Area: Applications->Everything Else

Keywords: Implicit Neural Representation, Untrained Neural Network, Radar Sensing, Super-resolution, Physics-informed Neural Network

Submission Number: 15009