How Information Theory Helps Sensing Design
Information Theory (IT) originated in the realm of telecommunications, focusing on channel optimization and encoding. In recent decades, IT has been applied to understanding fundamental concepts in various other fields such as cognitive neuroscience, biology, and dimensionality reduction in machine learning. This short video will present how IT can help us understand fundamental aspects of measurement, such as resolution and signal chain optimization. I will also demonstrate how some results can lead to practical applications, such as the optimal choice of quantization resolution in A/D conversion for a noisy interface. The goal is for this framework to simplify design challenges in more complex contexts.
Instructor
Marco Tartagni
Marco Tartagni obtained his M.S. in EE and a Ph.D. in EECS from the University of Bologna, Italy. In 1992, he joined the EE Department at Caltech in Pasadena, CA, initially as a visiting student and later in 1994 as a research fellow. During his time there, he focused on various aspects of analog VLSI for image processing. Since 1995, he has been a faculty member in the EE Department at the University of Bologna, where he currently holds the position of Full Professor. From 1996 to 2001, he was a team leader in the collaborative laboratory between STMicroelectronics and the University of Bologna. His work during this period involved the development of intelligent CMOS sensors, including CMOS cameras and the world's first prototype of a silicon-based fingerprint capacitive sensor. He is currently focusing his research on sensor design theory and Machine Learning assisted spectral sensing. Marco Tartagni is a co-recipient of the IEEE Van Vessem Outstanding Paper Award, which he received at the 2004 ISSCC Conference. He has also taken on the role of European coordinator for the FP6 Receptronics project in the field of Nanoelectronics. Furthermore, he has served on the scientific committees of the IEEE Custom Integrated Circuit Conference (CICC) from 2017 to 2021 and the IEEE International Electron Device Meeting (IEDM) in 2023.
Publication Year: 2024