Affiliation:
1. Department of Computer and Network Engineering, The University of Electro-Communications, 1-5-1 Chofugaoka, Chofu 182-8585, Tokyo, Japan
Abstract
The utilization of databases such as IoT has progressed, and understanding how to protect the privacy of data is an important issue. As pioneering work, in 1983, Yamamoto assumed the source (database), which consists of public information and private information, and found theoretical limits (first-order rate analysis) among the coding rate, utility and privacy for the decoder in two special cases. In this paper, we consider a more general case based on the work by Shinohara and Yagi in 2022. Introducing a measure of privacy for the encoder, we investigate the following two problems: The first problem is the first-order rate analysis among the coding rate, utility, privacy for the decoder, and privacy for the encoder, in which utility is measured by the expected distortion or the excess-distortion probability. The second task is establishing the strong converse theorem for utility–privacy trade-offs, in which utility is measured by the excess-distortion probability. These results may lead to a more refined analysis such as the second-order rate analysis.
Funder
Japan Society for the Promotion of Science
Subject
General Physics and Astronomy
Reference31 articles.
1. A source coding problem for sources with additional outputs to keep secret from the receiver or wiretappers;Yamamoto;IEEE Trans. Inf. Theory,1983
2. Utility–Privacy tradeoff in databases: An information-theoretic approach;Sankar;IEEE Trans. Inf. Forensics Secur.,2013
3. Shinohara, N., and Yagi, H. (2022, January 17–19). Unified expression of utility–privacy trade-off in privacy-constrained source coding. Proceedings of the 2022 International Symposium on Information Theory and Its Applications (ISITA2022), Tsukuba, Japan.
4. Ingber, A., and Kochman, Y. (2011, January 29–31). The dispersion of lossy source coding. Proceedings of the 2011 Data Compression Conference, Snowbird, UT, USA.
5. Fixed length lossy compression in the finite blocklength regime: Discrete memoryless sources;Kostina;IEEE Trans. Inf. Theory,2012