Application of non-linear electronic circuits to the design of ultra-low power event-based wake-up sensors

Published : 24 April 2017

The PhD work aims at contributing to the development of the Internet Of Things by studying innovative architectures of ultra-low power sensor interfaces able to wake-up on an event. The thesis work will take place in LGECA laboratory of CEA-LETI. It will consist in the study of the architecture, its silicon implementation in state-of-the art silicon technologies and its full characterization.

Most of sensors applications consist in the observation and detection of sparse events. In order to decrease the sensor power consumption, it is highly desirable to let it in sleep mode most of the time. Between the different wake-up strategies, the event-driven corresponds to an optimum in term of power consumption, provided that Ultra-Low Power solutions of detection and signal classification are available. Classical architectures of sensor are based on highly linear and accurate analog processing circuits which are power hungry and thus not good candidates for event-driven wake-up sensors. The thesis work will consist in the study of innovative architectures using non-linear functions, able to process information in the time domain in order to take full advantage of the ultra-low power possibilities of state-of-the-art Silicon technologies (FDSOI 28nm, 14nm). Targeted applications will be those which make use of vibrations, impacts and drops. The proposed solution will be integrated in advanced silicon technologies and will be associated to a CEA-leti micro-accelerometer. It will be fully characterized in laboratory and will be assessed on a chosen application.

The thesis work will last 3 years in the LGECA laboratory of CEA-LETI, which is leader in the conception of fully integrated sensor interfaces. The PhD student must have a deep knowledge in the design of mixed integrated circuit. A good knowledge of signal theory will be appreciated. The thesis work will be shared out between bibliography (10%), architecture (20%), design (30%), characterization (20%) and redaction of publications and dissertation (20%).

More information