Visit UCR Return to Campus website - Take the COVID Screening Check survey

Breadcrumb

PhD Defense of Rajasekhar Anguluri

Security of Interconnected Stochastic Dynamical Systems
10AM in Winston Chung Hall 202 –

The Department of Mechanical Engineering presents:
The Ph.D. Dissertation Defense of
Rajasekhar Anguluri

 

Monday, November 18, 2019,
10AM in Winston Chung Hall 202

Security of Interconnected Stochastic Dynamical
Systems


Doctor of Philosophy, Graduate Program in Mechanical Engineering
University of California, Riverside, November 2019
Dr. Fabio Pasqualetti, Chairperson

In this thesis we address problems concerning the security of interconnected systems that are subject to random (stochastic) disturbances. Our contribution is twofold. First, we investigate whether, and to what extent, coordination among different subsystems and knowledge of the global system dynamics is necessary to detect attacks in interconnected systems. We consider centralized and decentralized detectors, which differ primarily in their knowledge of the system model and characterize the performance of the two detectors and show that, depending on the system and attack parameters, each of the detectors can outperform the other. Hence, it may be possible for the decentralized detector to outperform its centralized counterpart, despite having less information about the system dynamics. We provide an explanation for this counter-intuitive result and illustrate our results through simulations. Second, we study an attack design problem for interconnected systems where the attacker compromises a subsystem at each time, based on a pre-computed probabilistic rule. The goal of the attacker is to degrade the system performance, which is measured based on a quadratic function of the system state, while remaining undetected from a centralized detector. We show that the ability to selectively compromise different subsystems over time increases the severity of the attacks with respect to compromising a fixed subsystem at each time.