Random Telegraph Noise in MEMS Gyroscopes: Modeling and Design Flow Enhancements Toward Autonomous Driving Sensor Design

This article addresses key limiting factors in the development of inertial navigation systems (INSs) based on MEMS inertial sensors, focusing on the long-term impact of low-frequency noise on the stability and accuracy of MEMS gyroscopes used in inertial measurement units (IMUs). We highlight the Allan deviation (ADEV) method as a systematic tool for identifying and quantifying the effects of technological defects that compromise sensor performance. In addition, we propose and calibrate a noise generation technique, based on measurements, which, in conjunction with the ADEV plot, identifies random telegraph noise (RTN) as a significant threat to commercial gyroscopes. Our findings aim to improve the reliability, performance, and yield of MEMS sensors by incorporating the noise sources and the design of the gyroscope into emulation platforms, overcoming the limitation posed by common CAD tools in evaluating very low-frequency noise.