April 13, 2018, Webb 1100

The multiple components of Cyber-physical systems are often interconnected by a shared communication network of limited bandwidth. One way to model this constraint is to assume that, at any time instant, a single packet can be reliably transmitted over the network to its destination. Therefore, the system designer must come up with a rules/algorithms that allocate this shared communication resource among multiple transmitting nodes. In this talk, we will discuss a new class of remote estimation problems where the communication resources are allocated dynamically based on the observations at the sensors, rather than purely on their statistical description. We will start with a decentralized remote estimation system where multiple sensors observe distinct random variables and communicate their measurements to the destination over a collision channel. Using an approach based on team-decision theory, we will establish the existence of a globally optimal solution where each sensor uses a threshold policy. Then, we will discuss the optimal design of a centralized collision avoidance policy by means of sensor scheduling. In this case, we will obtain person-by-person optimal policies for the scheduling of sensors making correlated Gaussian observations. Finally, we will show how our theoretical results can be applied to design scheduling policies where the joint probability density of the observations is unknown.

Marcos Muller Vasconcelos received his PhD degree in Electrical and Computer Engineering from the University of Maryland, College Park in 2016. He is currently a postdoctoral research associate with the Department of Electrical Engineering at the University of Southern California. His research interests include: Estimation and Control over Communication Networks, Optimization Theory and Systems Biology.