Observation of Alpha-Stable Noise in the Laser Gyroscope Data

As a typical non-Gaussian signal model, alpha-stable distribution is fit for describing the noise with heavy tail and impulsive characters. By the statistical analysis of the laser gyroscope data sampled with long-time static experiments, the random noise of the laser gyroscope shows a non-Gaussian property and can be modeled with alpha-stable noise instead of five major kinds of noises with the traditional Allan variance method. Power spectral analysis, Q-Q plot test, quantile method, and characteristic function method are applied to statistically analyze the random noise of the laser gyroscope and to estimate the value of stability index. Furthermore, the sampled data of the laser gyroscope are divided into 140 groups to calculate the characteristic index. In addition, the results stay between 1.96 and 1.985, which verify that random noise of the laser gyroscope includes the alpha-stable noise.