Fully Analog Resonant Recurrent Neural Network via Metacircuit

2026-04-19

AI Summary

This work addresses the challenge of accurately mapping fully analog recurrent neural networks onto physical hardware by proposing the first scalable, fully analog Resonant Recurrent Neural Network (R²NN) architecture. Built upon a metamaterial-inspired circuit design, R²NN leverages electromechanical analogies to directly encode network parameters as circuit elements. It integrates globally coupled resistors and locally resonant units—both jointly trainable—to form frequency-selective current pathways that extract discriminative spectral features. Operating entirely in the analog domain without analog-to-digital conversion, R²NN enables real-time processing of raw time-series signals. Its cross-domain versatility is demonstrated across tactile sensing, speech recognition, and condition monitoring tasks, substantially overcoming the energy efficiency and latency limitations inherent in conventional digital systems for edge intelligence.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Analogy
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Physical neural networks offer a transformative route to edge intelligence, providing superior inference speed and energy efficiency compared to conventional digital architectures. However, realizing scalable, end-to-end, fully analog recurrent neural networks for temporal information processing remains challenging due to the difficulty of faithfully mapping trained network models onto physical hardware. Here we present a fully analog resonant recurrent neural network (R$^2$NN) implemented via a metacircuit architecture composed of coupled electrical local resonators. A reformulated mechanical-electrical analogy establishes a direct mapping between the R$^2$NN model and metacircuit elements, enabling accurate physical implementation of trained neural network parameters. By integrating jointly trainable global resistive coupling and local resonances, which generate effective frequency-dependent negative resistances, the architecture shapes an impedance landscape that steers currents along frequency-selective pathways. This mechanism enables direct extraction of discriminative spectral features, facilitating real-time temporal classification of raw analog inputs while bypassing analog-to-digital conversion. We demonstrate the cross-domain versatility of this framework using integrated hardware for tactile perception, speech recognition, and condition monitoring. This work establishes a scalable, fully analog paradigm for intelligent temporal processing and paves the way for low-latency, resource-efficient physical neural hardware for edge intelligence.

Problem

• analog recurrent neural network
• physical neural networks
• temporal information processing
• hardware implementation
• edge intelligence

Innovation

• analog neural network
• resonant recurrent neural network
• metacircuit
• physical neural hardware
• frequency-selective processing