Markov Noise in Atomic Spin Gyroscopes: Analysis and Suppression Based on Allan Deviation

Atomic spin gyroscopes (ASGs) based on spin-exchange relaxation-free (SERF) comagnetometer have ultrahigh rotation sensitivity and low gyro drift. However, the noise characteristics of ASGs have not been comprehensively studied. In this article, the primary noise in ASGs, which weakens the gyro's long-term stability, is analyzed and identified with Allan deviation. We discovered, for the first time, that the slow convergence rate of the spin-coupled ensemble brings a long correlation time of the gyro output, and the Markov noise introduced by this effect is the primary noise that impacts the long-term stability of ASGs. Allan deviation is used to assess the characteristics of Markov noise and its impact on long-term stability. The correlation time is adjusted by changing the pumping power density to suppress the Markov noise. Comparing the Allan deviations under different correlation time, it is concluded that the Markov noise caused by long correlation time restricts the improvement of the long-term stability of ASGs. Moreover, Markov noise at low frequencies is suppressed by increasing the pumped optical power density, enhancing the rotation sensitivity of ASGs. A relatively low gyro drift $5.76\times 10 ^{-3}{}^{\circ} $ /h@100 s is acquired after the suppression of the Markov noise. Different from the traditional amplitude-frequency response analysis, this article focuses on the time correlation of noise, which provides a new idea for the investigation of ASG's noise characteristics.