Designing large scale cyber-physical systems (CPS, e.g. power-grid, Internet of Things) involves exploring a vast design space defined by competing metrics. The questions and tradeoffs that arise include: should control be centralized, distributed or decentralized (global decision making vs. latency); and how much actuation, sensing and communication is required (architectural cost vs. performance). The traditional ad-hoc approach to answering these questions becomes less viable as the scale, complexity and importance of CPS grow. With increasingly dynamic CPS operating under uncertain conditions, architectural decisions made early on in the design process will be as crucial to the overall performance of the system as the control policies themselves.
In this talk, we argue that the appropriate means of exploring this design space is through the co-design of components of the control architecture. We focus on the specific co-design problem of jointly synthesizing an optimal distributed feedback controller and its required architecture (i.e., placement of sensors, actuators, and communication links between them). We show that this challenge of co-design can be framed as one of seeking structured solutions to a linear inverse problem and formulate the Regularization for Design framework, in which we augment variational formulations of controller synthesis problems with convex penalty functions that induce a desired (sparse) controller architecture. The resulting regularized formulations are convex optimization problems that can be solved efficiently, and provide a unified computationally tractable approach for the simultaneous co-design of a structured optimal controller and the actuation, sensing and communication architecture required to implement it. We further show that our approach identifies a nominally structured controller under a suitable condition on a “signal-to-noise” type ratio.
We conclude our talk with a discussion of how the previous results fit within our broader theory of architecture, which integrates layering, optimization, dynamics and control into a unified framework, and summarize promising preliminary results in applying these tools to software defined networking enabled traffic control in wide area networks — these include simulation, emulation and experimental results. Time permitting, we will also make connections to recent neuroscience experiments.
Nikolai Matni is a postdoctoral scholar in Computing and Mathematical Sciences at the California Institute of Technology. He received the B.A.Sc. and M.A.Sc. in Electrical Engineering from the University of British Columbia, and the Ph.D. in Control and Dynamical Systems from the California Institute of Technology. His research interests broadly encompass the use of layering, dynamics, control and optimization in the design and analysis of complex cyber-physical systems; current application areas include software defined networking and sensorimotor control. His Ph.D. work focused on foundational theory of distributed optimal control, and in particular on controller synthesis, architecture design and system identification. He was awarded the 2013 IEEE CDC Best Student-Paper Award.