Back to TUM-PREP Projects overview

Project Overview

Project Code: CIT 20

Project name:

Privacy in multi-user systems

TUM Department:

CIT - Electrical and Computer Engineering

TUM Chair / Institute:

Professorship of Coding and Cryptography

Research area:

Privacy, Coding Theory, Information Theory

Student background:

Computer ScienceComputer Science/ InformaticsElectrical EngineeringMathematics

Further disciplines:

Participation also possible online only:

Planned project location:

N4 building, Theresienstraße 90, 80992 München

Project Supervisor - Contact Details


Title:

Given name:

Luis

Family name:

Maßny

E-mail:

luis.massny@tum.de

Phone:

+49 (89) 289 - 23423

Additional Project Supervisor - Contact Details


Title:

Given name:

Maximilian

Family name:

Egger

E-mail:

maximilian.egger@tum.de

Phone:

Additional Project Supervisor - Contact Details


Title:

Dr.

Given name:

Rawad

Family name:

Bitar

E-mail:

rawad.bitar1@gmail.com

Phone:

Project Description


Project description:

Many applications, such as data analytics and federated learning, rely on the availability of high-quality data. When individuals participate in or contribute to such systems, the privacy of their sensitive information is a concern.

A plethora of statistical and information-theoretical measures have been proposed [1,2] to rigorously define privacy as information leakage, for instance, mutual information or differential privacy leakage. In essence, ensuring privacy comes with the need to make different realizations of private information confusable by the revealed information. It is well-known that this requires either an increase in communication or a loss in utility, jeopardizing the performance or robustness of an information system. Therefore, the trade-off between privacy and its cost requires careful consideration.

At the Professorship for Coding and Cryptography, we work on the development and analysis of private communication protocols. Our toolbox contains methods from coding theory as well as probability theory. This research project aims to understand several privacy notions, applying these to analyze the privacy of existing protocols theoretically and improving or developing novel privacy-preserving protocols. Our focus lies on privacy in multi-user systems, for example, private federated learning [3] or privacy in social networks [4]. The results and the effect on the system performance can be verified by implementing the protocols in a simulated multi-user environment, e.g., a federated learning system. However, the emphasis lies on theoretic research.

As our research environment evolves rapidly, the exact focus will be defined at the beginning of the research project.

[1] M. Bloch et al., “An Overview of Information-Theoretic Security and Privacy: Metrics, Limits and Applications,” IEEE Journal on Selected Areas in Information Theory, Mar. 2021.
[2] I. Wagner and D. Eckhoff, “Technical privacy metrics: a systematic survey,” ACM Computing Surveys (CSUR), 2018.
[3] K. Bonawitz et al., “Practical Secure Aggregation for Privacy-Preserving Machine Learning,” in Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, Oct. 2017.
[4] C. Naim, F. Ye, and S. El Rouayheb, “On the Privacy of Social Networks with Personal Privacy Choices,” in 2023 IEEE International Symposium on Information Theory (ISIT), Jun. 2023.

Working hours per week planned:

40

Prerequisites


Required study level minimum (at time of TUM PREP project start):

3 years of bachelor studies completed

Subject related:

- Strong background in statistics and probability theory
- Basic knowledge of information theory and coding theory

Other:

- Programming with Python

  • Keine Stichwörter