Meta-training of diffractive meta-neural networks for super-resolution direction of arrival estimation

2025-09-07

AI Summary

Diffractive neural networks face challenges in jointly achieving large-scale metasurface modeling and high-precision training, while failing to exploit multidimensional electromagnetic field encoding for super-resolution direction-of-arrival (DoA) estimation. Method: This paper proposes the Diffractive Meta-Neural Network (DMNN) architecture: (i) it introduces a pretrained compact model to characterize meta-atom electromagnetic responses, integrated with gradient-driven meta-training for inverse design; (ii) it combines multi-frequency–multi-polarization channels, optical superoscillation, and spectral encoding to surpass the diffraction limit; and (iii) it employs a multilayer diffractive structure, angular-domain interleaved frequency multiplexing, and a lightweight electronic post-processing network for all-optical multi-task computation. Contribution/Results: Experimentally demonstrated at 27–31 GHz, DMNN achieves 0.5° angular resolution—approximately a 7×突破 beyond the diffraction limit—with a mean absolute error of 0.048° for dual-target DoA estimation and an inference throughput of 1917 Hz—improving by one order of magnitude over prior approaches.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Search and Optimization: Metareasoning and Metaheuristics
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Diffractive neural networks leverage the high-dimensional characteristics of electromagnetic (EM) fields for high-throughput computing. However, the existing architectures face challenges in integrating large-scale multidimensional metasurfaces with precise network training and haven't utilized multidimensional EM field coding scheme for super-resolution sensing. Here, we propose diffractive meta-neural networks (DMNNs) for accurate EM field modulation through metasurfaces, which enable multidimensional multiplexing and coding for multi-task learning and high-throughput super-resolution direction of arrival estimation. DMNN integrates pre-trained mini-metanets to characterize the amplitude and phase responses of meta-atoms across different polarizations and frequencies, with structure parameters inversely designed using the gradient-based meta-training. For wide-field super-resolution angle estimation, the system simultaneously resolves azimuthal and elevational angles through x and y-polarization channels, while the interleaving of frequency-multiplexed angular intervals generates spectral-encoded optical super-oscillations to achieve full-angle high-resolution estimation. Post-processing lightweight electronic neural networks further enhance the performance. Experimental results validate that a three-layer DMNN operating at 27 GHz, 29 GHz, and 31 GHz achieves $sim7 imes$ Rayleigh diffraction-limited angular resolution (0.5$^circ$), a mean absolute error of 0.048$^circ$ for two incoherent targets within a $pm 11.5^circ$ field of view, and an angular estimation throughput an order of magnitude higher (1917) than that of existing methods. The proposed architecture advances high-dimensional photonic computing systems by utilizing inherent high-parallelism and all-optical coding methods for ultra-high-resolution, high-throughput applications.

Problem

• Super-resolution direction of arrival estimation using diffractive meta-neural networks
• Integrating large-scale metasurfaces with precise network training challenges
• Multidimensional electromagnetic field coding for enhanced angular resolution

Innovation

• Meta-training diffractive networks for super-resolution
• Multidimensional EM field coding via metasurfaces
• Gradient-based inverse design for meta-atoms