Decentralized control concerns the regulation of a process by many interconnected agents each of which may have only partial information about the state of the system. Decentralized systems are a mainstay of computer science where many of the tools for designing and analyzing them originated. However, such systems often exhibit non-obvious dynamics as they become complex or are required to interact with physical processes. Thus, researchers are increasingly turning to feedback control methods for their design and analysis. In this lecture, I will introduce the field by considering the commonalities and differences of several examples, including: multi-vehicle systems, programmable self-assembling systems, self-stabilizing protocols, cyclic phase regulation and Internet congestion control.
Copy of presentation [pdf]
This seminar is being presented as part of CDS 101, Principles of Feedback and Control. If you would like to be added to the mailing list to receive future CDS 101 seminar announcements, you can subscribe to the cds101-announce mailing list.