Practical Approaches to Allan Deviation Analysis of Low-Cost MEMS Inertial Sensors

Abstract

This paper provides explanations and best practices on how to measure and interpret Allan deviation specifically for low-cost MEMS inertial sensors. It focuses on the reliable extraction of the two most important noise terms angle/velocity random walk (ARW/VRW) and bias instability (BIS) as well as their relevance in application. The understanding of Allan deviation as the stochastic, lowest-possible limit for sensor performance in the presence of much larger deterministic sensor errors is substantiated. Concrete guidelines for the measurement setup are provided. Lastly, typical phenomena in MEMS inertial sensors like sinusoidal signals or random telegraph noise (RTN) are presented in a holistic analysis, which includes Allan deviation, spectral density and low-pass filtered time series plots.