Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Informational bottlenecks in control systems

The trends set by the Internet of Things (IoT) are pushing us to understand the interactions between communication, computation and control. This talk will discuss simple models that can provide insights on the "informational" limits on the control of high-performance systems when the controller does not know the precise values of the model parameters.

The first part of the talk will focus on control over multiplicative noise actuation channels, and will develop a new notion of "control capacity," as the fundamental limit on the rate at which an unreliable actuator can reduce system uncertainty and thus control a system. This idea of control capacity is motivated by the traditional notion of communication capacity in information theory. Further, control capacity can be thought of as partially generalizing the uncertainty threshold principle from control.

The information-theoretic perspective that allows us to understand (and in some examples explicitly compute) the value of side-information the controller may receive (e.g. extra sensor measurements) in the context of control systems. Some of the calculations are inspired by the well-studied dropped control problem of stabilizing a system over a Bernoulli actuation channel.

If time permits, I will also discuss the stabilizability limits on systems with multiplicative observation noise. The converse bounds here use a non-standard approach to get around the analytical challenges posed by multiplicative noise.

Type of Seminar:
IfA Seminar
Dr. Gireeja Ranade
Microsoft Research, Redmond
Jul 20, 2016   14:15h

Contact Person:

Sabrina Baumann
No downloadable files available.
Biographical Sketch:
Gireeja Ranade is a postdoctoral researcher at Microsoft Research, Redmond. Before this she was a lecturer in EECS at UC Berkeley working on designing and teaching the pilot version of the new lower-division EECS classes (16AB). She received an MS and PhD in EECS from UC Berkeley and an SB in EECS from MIT. She has worked on topics in brain-machine interfaces, information theory, control theory, wireless communications and crowdsourcing.