Suppression of Quantum Sensor Noise using Kalman Filter for Improved Sensitivity of Single-Beam Atomic Magnetometers

Single-beam atomic magnetometers herald a new era of high-precision magnetic field sensing, with applications spanning fundamental physics to biomagnetism. Nevertheless, their utility is often curtailed by quantum sensor noise, encompassing both technical and quantum-mechanical noise. This research delves into the potential of the Kalman filter as a tool to subdue quantum sensor noise, thereby augmenting the sensitivity of single-beam atomic magnetometers. Quantum-mechanical noise is integrated into the system model as the process noise and measurement noise, and a discrete Kalman filter equipped with a time delay variable is employed. The findings reveal that the time delay variable significantly influences temporal signal tracing, while the discrete Kalman filter enhances sensitivity performance in frequency domain analysis, bypassing the typical sensitivity and time resolution trade-off encountered in coherent sensing strategies. Partial-knowledge signal scenarios are also taken into account, wherein a polynomial model is proposed to expansively render the discrete Kalman filter more relevant and adaptable to real-world situations. Collectively, through experimentation involving sine-like, non-Gaussian, and medical magnetocardiography (MCG) signals, our results underscore the promising potential of the Kalman filter in enhancing the sensitivity of atomic magnetometers for practical sensing applications.