© 2026 Aqora Quantum S.A.S.

Terms Privacy Legal Notice

Machine Learning Science and Technology

Application of kernel principal component analysis for optical vector atomic magnetometry

James McKelvy, Irina V. Novikova, Eugeniy E. Mikhailov, Mario A. Maldonado, I. Fan, Yang Li, Ying-Ju Wang, John E. Kitching, Andrey B. Matsko

Published Nov 24, 2023• 9 citations

Abstract

Vector atomic magnetometers that incorporate electromagnetically induced transparency (EIT) allow for precision measurements of magnetic fields that are sensitive to the directionality of the observed field by virtue of fundamental physics. However, a practical methodology of accurately recovering the longitudinal angle of the local field through observations of EIT spectra has not been established. In this work, we address this problem of angle determination with an unsupervised machine learning algorithm utilizing nonlinear dimensionality reduction. The proposed algorithm was developed to interface with spectroscopic measurements from an EIT-based atomic rubidium magnetometer and uses kernel principal component analysis (KPCA) as an unsupervised feature extraction tool. The resulting KPCA features allow each EIT spectrum measurement to be represented by a single coordinate in a new reduced dimensional feature space, thereby streamlining the process of angle determination. A supervised support vector regression (SVR) machine was implemented to model the resulting relationship between the KPCA projections and field direction. If the magnetometer is configured so that the azimuthal angle of the field is defined with a polarization lock, the KPCA-SVR algorithm is capable of predicting the longitudinal angle of the local magnetic field within 1 degree of accuracy and the magnitude of the absolute field with a resolution of 70 nT. The combined scalar and angular sensitivity of this method make the KPCA-enabled EIT magnetometer competitive with conventional vector magnetometry methods.

DOI: 10.1088/2632-2153/ad0fa4

Related papers

On the Development of an Optical Rubidium Vector Atomic Magnetometer

James McKelvy, Irina V. Novikova, Eugeniy E. Mikhailov +6 more

The precise measurement of magnetic fields is a fundamental tool of remote sensing. However, accurately measuring the direction of magnetic fields is challenging with atomic magnetometers. In this work, we discuss progress on the development of an all-optical vector atomic magnetometer that uses electromagnetically induced transparency (EIT) on87Rb vapor. The magnitude of the magnetic field is evaluated by measuring the separation of the laser transmission peaks of the Zeeman-resolved EIT spectra, and the direction is measured by evaluating the relative peak contrast using a lightweight machine learning algorithm involving principal component analysis (PCA). We have demonstrated a scalar sensitivity of $ < 10pT/\sqrt {Hz} $ in the 1–100 Hz band and an angular accuracy of < 1°. This approach for vector magnetometry does not require an array of sensors or calibration coils, ultimately setting the groundwork for the design of a new magnetometer that is lighter and more accurate than conventional vector magnetometer configurations.

IEEE International Geoscience and Remote Sensing Symposium (IGARSS)•Jul 16, 2023•2 citations

Sensitivity of a vector atomic magnetometer based on electromagnetically induced transparency

Mario Alberto González Maldonado, Owen Rollins, Alex Toyryla +8 more

We present a realization of a magnetic sensor based on electromagnetically induced transparency (EIT) resonances observed in hot Rb vapor using lin$\parallel$lin polarized dichromatic light and evaluate scalar and vector capabilities of the sensor for measuring Earth-like magnetic fields. We demonstrate scalar measurement sensitivity of 2 pT/Hz^1/2 in the 1-100~Hz spectral frequency band using a cm^3 Rb vapor cell, significantly improving the performance for such a configuration if compared with earlier measurements of large magnetic fields. By using a single linearly polarized dichromatic optical field, we are also able to determine the direction of the magnetic field with respect to the light propagation direction and polarization, taking advantage of the symmetries of the interaction scheme. We accomplish that by combining the polarization-sensitive transmission measurements and sparse sensing machine learning techniques. A path for further improvement of the sensitivity and elimination of systematic effects, such as heading error, is discussed.

Optics Express•Jun 12, 2024•14 citations

Machine learning enhanced microwave electrometry in many-body Rydberg atoms

Bin-Bin Wang, Dong Yan, Jin‐Hui Wu

Precision microwave electrometry using Rydberg atoms is limited by many-body interactions at moderate or not too low excitation densities, which may result in distorted probe spectra of electromagnetically induced transparency (EIT) as predicted by mean-field superatom simulations. This nonlinearity fundamentally restricts measurement accuracy and renders the linear response formula of Autler-Townes splitting to microwave (MW) amplitude inapplicable. Here we present a machine learning (ML) framework assisted by principal component analysis (PCA) that can well overcome this difficulty by mapping distorted noisy EIT spectra directly to MW electric fields. Numerical results show that a basic PCA-assisted regression model reduces the minimal detectable MW field by a few folds as compared to both linear response formula and nonlinear least squares fitting in the interaction-dominant regime. It is of more interest that we can employ a four-class ensemble ML scheme to further reduce the minimal detectable MW field by two orders of magnitude to approach its noise-dependent physical limit and hence largely improve the measurement sensitivity. Our work transforms detrimental Rydberg interactions into beneficial spectral features, providing a practical route toward optimal sensitivity for MW quantum sensing in the nonlinear regime. Rydberg atom-based microwave electrometry is hindered by many-body interactions, distorting probe spectra and limiting measurement accuracy. Here, the authors employ a machine learning framework with principal component analysis to map distorted spectra to microwave fields, significantly enhancing sensitivity and transforming interaction challenges into advantageous spectral features for quantum sensing.

Communications Physics•Aug 15, 2026•0 citations

Sensitivity of a vector atomic magnetometer based on electromagnetically induced transparency

Irina V. Novikova, Eugeniy E. Mikhailov, John E. Kitching +8 more

We present a realization of a magnetic sensor based on electromagnetically induced transparency (EIT) resonances observed in hot Rb vapor using lin$\parallel$lin polarized dichromatic light and evaluate scalar and vector capabilities of the sensor for measuring Earth-like magnetic fields. We demonstrate scalar measurement sensitivity of 2 pT/Hz^1/2 in the 1-100~Hz spectral frequency band using a cm^3 Rb vapor cell, significantly improving the performance for such a configuration if compared with earlier measurements of large magnetic fields. By using a single linearly polarized dichromatic optical field, we are also able to determine the direction of the magnetic field with respect to the light propagation direction and polarization, taking advantage of the symmetries of the interaction scheme. We accomplish that by combining the polarization-sensitive transmission measurements and sparse sensing machine learning techniques. A path for further improvement of the sensitivity and elimination of systematic effects, such as heading error, is discussed.

•May 6, 2024•0 citations

Sensitivity of a vector atomic magnetometer based on electromagnetically induced transparency

Irina V. Novikova, Eugeniy E. Mikhailov, John E. Kitching +8 more

We present a realization of a magnetic sensor based on electromagnetically induced transparency (EIT) resonances observed in hot Rb vapor using lin$\parallel$lin polarized dichromatic light and evaluate scalar and vector capabilities of the sensor for measuring Earth-like magnetic fields. We demonstrate scalar measurement sensitivity of 2 pT/Hz^1/2 in the 1-100~Hz spectral frequency band using a cm^3 Rb vapor cell, significantly improving the performance for such a configuration if compared with earlier measurements of large magnetic fields. By using a single linearly polarized dichromatic optical field, we are also able to determine the direction of the magnetic field with respect to the light propagation direction and polarization, taking advantage of the symmetries of the interaction scheme. We accomplish that by combining the polarization-sensitive transmission measurements and sparse sensing machine learning techniques. A path for further improvement of the sensitivity and elimination of systematic effects, such as heading error, is discussed.

•Jun 14, 2024•0 citations

Vector atomic magnetometer based on EIT in Rb vapor

Irina V. Novikova, Eugeniy E. Mikhailov, Alex Toyryla +6 more

By analyzing the relative frequencies and amplitude of multiple electromagnetically induced transparency resonances in external magnetic field, we achieve scalar magnetometer sensitivity <5pT/Hz, and can determine the B-field direction with angular sensitivity < 1° using machine learning and optical lock techniques in a 100 mm3 atomic vapor cell.

•Jan 1, 2023•2 citations

Vector atomic magnetometers that incorporate electromagnetically induced transparency (EIT) allow for precision measurements of magnetic fields that are sensitive to the directionality of the observed field by virtue of fundamental physics. However, a practical methodology of accurately recovering the longitudinal angle of the local field through observations of EIT spectra has not been established. In this work, we address this problem of angle determination with an unsupervised machine learning algorithm utilizing nonlinear dimensionality reduction. The proposed algorithm was developed to interface with spectroscopic measurements from an EIT-based atomic rubidium magnetometer and uses kernel principal component analysis (KPCA) as an unsupervised feature extraction tool. The resulting KPCA features allow each EIT spectrum measurement to be represented by a single coordinate in a new reduced dimensional feature space, thereby streamlining the process of angle determination. A supervised support vector regression (SVR) machine was implemented to model the resulting relationship between the KPCA projections and field direction. If the magnetometer is configured so that the azimuthal angle of the field is defined with a polarization lock, the KPCA-SVR algorithm is capable of predicting the longitudinal angle of the local magnetic field within 1 degree of accuracy and the magnitude of the absolute field with a resolution of 70 nT. The combined scalar and angular sensitivity of this method make the KPCA-enabled EIT magnetometer competitive with conventional vector magnetometry methods.

The precise measurement of magnetic fields is a fundamental tool of remote sensing. However, accurately measuring the direction of magnetic fields is challenging with atomic magnetometers. In this work, we discuss progress on the development of an all-optical vector atomic magnetometer that uses electromagnetically induced transparency (EIT) on87Rb vapor. The magnitude of the magnetic field is evaluated by measuring the separation of the laser transmission peaks of the Zeeman-resolved EIT spectra, and the direction is measured by evaluating the relative peak contrast using a lightweight machine learning algorithm involving principal component analysis (PCA). We have demonstrated a scalar sensitivity of $ < 10pT/\sqrt {Hz} $ in the 1–100 Hz band and an angular accuracy of < 1°. This approach for vector magnetometry does not require an array of sensors or calibration coils, ultimately setting the groundwork for the design of a new magnetometer that is lighter and more accurate than conventional vector magnetometer configurations.

We present a realization of a magnetic sensor based on electromagnetically induced transparency (EIT) resonances observed in hot Rb vapor using lin

\parallel

lin polarized dichromatic light and evaluate scalar and vector capabilities of the sensor for measuring Earth-like magnetic fields. We demonstrate scalar measurement sensitivity of 2 pT/Hz^1/2 in the 1-100~Hz spectral frequency band using a cm^3 Rb vapor cell, significantly improving the performance for such a configuration if compared with earlier measurements of large magnetic fields. By using a single linearly polarized dichromatic optical field, we are also able to determine the direction of the magnetic field with respect to the light propagation direction and polarization, taking advantage of the symmetries of the interaction scheme. We accomplish that by combining the polarization-sensitive transmission measurements and sparse sensing machine learning techniques. A path for further improvement of the sensitivity and elimination of systematic effects, such as heading error, is discussed.

Precision microwave electrometry using Rydberg atoms is limited by many-body interactions at moderate or not too low excitation densities, which may result in distorted probe spectra of electromagnetically induced transparency (EIT) as predicted by mean-field superatom simulations. This nonlinearity fundamentally restricts measurement accuracy and renders the linear response formula of Autler-Townes splitting to microwave (MW) amplitude inapplicable. Here we present a machine learning (ML) framework assisted by principal component analysis (PCA) that can well overcome this difficulty by mapping distorted noisy EIT spectra directly to MW electric fields. Numerical results show that a basic PCA-assisted regression model reduces the minimal detectable MW field by a few folds as compared to both linear response formula and nonlinear least squares fitting in the interaction-dominant regime. It is of more interest that we can employ a four-class ensemble ML scheme to further reduce the minimal detectable MW field by two orders of magnitude to approach its noise-dependent physical limit and hence largely improve the measurement sensitivity. Our work transforms detrimental Rydberg interactions into beneficial spectral features, providing a practical route toward optimal sensitivity for MW quantum sensing in the nonlinear regime. Rydberg atom-based microwave electrometry is hindered by many-body interactions, distorting probe spectra and limiting measurement accuracy. Here, the authors employ a machine learning framework with principal component analysis to map distorted spectra to microwave fields, significantly enhancing sensitivity and transforming interaction challenges into advantageous spectral features for quantum sensing.

We present a realization of a magnetic sensor based on electromagnetically induced transparency (EIT) resonances observed in hot Rb vapor using lin

\parallel

lin polarized dichromatic light and evaluate scalar and vector capabilities of the sensor for measuring Earth-like magnetic fields. We demonstrate scalar measurement sensitivity of 2 pT/Hz^1/2 in the 1-100~Hz spectral frequency band using a cm^3 Rb vapor cell, significantly improving the performance for such a configuration if compared with earlier measurements of large magnetic fields. By using a single linearly polarized dichromatic optical field, we are also able to determine the direction of the magnetic field with respect to the light propagation direction and polarization, taking advantage of the symmetries of the interaction scheme. We accomplish that by combining the polarization-sensitive transmission measurements and sparse sensing machine learning techniques. A path for further improvement of the sensitivity and elimination of systematic effects, such as heading error, is discussed.

We present a realization of a magnetic sensor based on electromagnetically induced transparency (EIT) resonances observed in hot Rb vapor using lin

\parallel

lin polarized dichromatic light and evaluate scalar and vector capabilities of the sensor for measuring Earth-like magnetic fields. We demonstrate scalar measurement sensitivity of 2 pT/Hz^1/2 in the 1-100~Hz spectral frequency band using a cm^3 Rb vapor cell, significantly improving the performance for such a configuration if compared with earlier measurements of large magnetic fields. By using a single linearly polarized dichromatic optical field, we are also able to determine the direction of the magnetic field with respect to the light propagation direction and polarization, taking advantage of the symmetries of the interaction scheme. We accomplish that by combining the polarization-sensitive transmission measurements and sparse sensing machine learning techniques. A path for further improvement of the sensitivity and elimination of systematic effects, such as heading error, is discussed.