Differential Privacy for Dynamic Data (SpringerBriefs in Control, Automation and Robotics)

Preface
Contents
Notation and Abbreviations
1 Defining Privacy-Preserving Data Analysis
	1.1 Motivation
	1.2 Privacy Attacks
	1.3 Differential Privacy
		1.3.1 Standard Definition
		1.3.2 Definition via Hypothesis Testing
		1.3.3 Basic Composition Theorem and Resilience to Post-processing
		1.3.4 Variations on Differential Privacy
	1.4 Alternative Approaches to Privacy-Preserving Data Analysis
		1.4.1 Statistical Disclosure Limitation, K-anonymity, and Information-Theoretic Approaches
		1.4.2 Privacy, Secrecy and the Role of Cryptography
	1.5 Privacy-Preserving Analysis for Dynamic Data
	References
2 Basic Differentially Private Mechanisms
	2.1 Introduction
	2.2 Laplace and Gaussian Mechanisms
	2.3 Input and Output Mechanisms: Basic Examples
		2.3.1 Adjacency Relations on Individual and Collective Data Streams
		2.3.2 Input Perturbation Mechanisms
		2.3.3 Output Perturbation Mechanisms
		2.3.4 Comparison of Input and Output Perturbation Mechanisms
	References
3 A Two-Stage Architecture for Differentially Private Filtering
	3.1 Introduction
	3.2 Design Methodology for Two-Stage Mechanisms
	3.3 Two-Block Approximation Architecture for LTI Filters
	3.4 Zero-Forcing Equalization Mechanism
		3.4.1 Single-Input Case
		3.4.2 Multiple-Input Case
	3.5 Application to Privacy-Preserving Estimation of Building Occupancy
	References
4 Differentially Private Filtering  for Stationary Stochastic Collective Signals
	4.1 Introduction
	4.2 Linear Minimum Mean Square (Wiener) Smoothing and Filtering
	4.3 Two-Stage Differentially Private LMMSE Estimators
		4.3.1 Lower Bound on MSE for Diagonal Pre-filters
		4.3.2 Pre- and Post-filter Design
	References
5 Differentially Private Kalman Filtering
	5.1 Introduction
	5.2 Differentially Private Kalman Filtering Problems
		5.2.1 Individual Data Streams
		5.2.2 Collective Data Streams
	5.3 Input Perturbation Mechanisms
	5.4 Output Perturbation Mechanisms for Time-Invariant Filters
		5.4.1 Filter Redesign for Stable Systems
		5.4.2 Unstable Systems
		5.4.3 Post-filtering
		5.4.4 Example
	5.5 Optimization of Two-Stage Kalman Filtering Mechanisms: Static First Stage Aggregation
	References
6 Differentially Private Nonlinear Observers
	6.1 Introduction
	6.2 Differentially Private Observer Design
	6.3 Input and Output Perturbation Mechanisms
	6.4 Contracting Systems
		6.4.1 Basic Results
		6.4.2 Effect of Disturbances on Contractive Systems
	6.5 Differentially Private Observers with Output Perturbation
	6.6 Application Examples
	References
7 Conclusion
Appendix A Discrete-Time Stochastic Signals
Appendix B Proof of the Basic Composition Theorem
Appendix  Series Editor Biographies