Core Concepts
The author introduces a novel approach to ensure differential privacy in tracking errors of nonlinear systems while maintaining performance guarantees through the use of funnel control and Ornstein-Uhlenbeck-type noise. The main thesis is to make tracking errors differentially private by adding bounded continuous noise to the performance funnel.
Abstract
In this content, the authors introduce a new framework for ensuring differential privacy in tracking errors of nonlinear systems. They utilize funnel control and Ornstein-Uhlenbeck-type noise to maintain performance guarantees while making tracking errors differentially private. The paper discusses the background on differential privacy, introduces a new framework for making tracking errors differentially private using a funnel controller, and demonstrates the application through simulations. The study also addresses the adjacency relation for funnel boundaries, query sensitivity, and univariate bounded Gaussian noise. Furthermore, it explores the Ornstein-Uhlenbeck type process for generating continuous bounded noise and its impact on differential privacy. Finally, simulation results are presented to validate the proposed framework.
Quotes
"We introduce a novel approach to make the tracking error of a class of nonlinear systems differentially private."
"We use funnel control to make the tracking error evolve within a performance funnel that is pre-specified by the user."