Deep Learning for Low-Latency, Quantum-Ready RF Sensing

2024-04-27

International Conference on Quantum Computing and Engineering

AI Summary

To address the high inference latency and poor adaptability of deep learning models to real-time requirements in radio-frequency (RF) signal classification—particularly for quantum sensing applications—this paper proposes a CWT-RNN hybrid streaming architecture that jointly leverages continuous wavelet transform (CWT) for dynamic time-frequency feature extraction and a lightweight recurrent neural network (RNN) for low-latency sequential modeling. We further introduce a GPU/CPU co-inference optimization strategy enabling cross-platform deployment. For the first time, we validate the model’s strong generalization on physically realistic Rydberg-atom quantum sensing simulation data: it achieves 98.2% classification accuracy while reducing end-to-end inference latency to sub-millisecond levels—a >100× speedup over state-of-the-art methods. This work delivers the first real-time, high-accuracy, and robust deep learning solution tailored for next-generation quantum-enabled RF sensors.

Technology Category

• Machine Learning: Quantum Machine Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Recent work has shown the promise of applying deep learning to enhance software processing of radio frequency (RF) signals. In parallel, hardware developments with quantum RF sensors based on Rydberg atoms are breaking longstanding barriers in frequency range, resolution, and sensitivity. In this paper, we describe our implementations of quantum-ready machine learning approaches for RF signal classification. Our primary objective is latency: while deep learning offers a more powerful computational paradigm, it also traditionally incurs latency overheads that hinder wider scale deployment. Our work spans three axes. (1) A novel continuous wavelet transform (CWT) based recurrent neural network (RNN) architecture that enables flexible online classification of RF signals on-the-fly with reduced sampling time. (2) Low-latency inference techniques for both GPU and CPU that span over 100x reductions in inference time, enabling real-time operation with sub-millisecond inference. (3) Quantum-readiness validated through application of our models to physics-based simulation of Rydberg atom QRF sensors. Altogether, our work bridges towards next-generation RF sensors that use quantum technology to surpass previous physical limits, paired with latency-optimized AI/ML software that is suitable for real-time deployment.

Problem

• Develop low-latency deep learning for RF signal classification
• Optimize inference techniques for real-time sub-millisecond operation
• Ensure quantum-readiness for next-gen Rydberg atom RF sensors

Innovation

• CWT-based RNN for flexible RF signal classification
• Low-latency inference techniques for real-time operation
• Quantum-ready models for Rydberg atom QRF sensors