Decentralized Generation of Multiple, Uncorrelatable Pseudonyms without Trusted Third Parties

Privacy Technologies and Policy

“…As stated in [24], when designing such a decentralised approach for pseudonym generation, we are mainly interested in fulfilling the following requirements: i) ease of use, ii) linking a pseudonym to its owning user should not be possible for any other than the user herself, unless it is explicitly permitted, iii) in cases that users may have multiple pseudonyms, it should not be possible to identify different pseudonyms as belonging to the same user, iv) injectivity, in terms that the pseudonym generation process should avoid duplicates, v) flexibility, i.e. it should be possible to add new pseudonyms to the user entities with minimal effort.…”

“…Therefore, the new pseudonymisation technique satisfies the properties described in [24] (implementation issues will be subsequently analysed), enriched with some additional properties that may be of high importance in specific scenarios, as discussed next. Actually, due to the property P3, the user A may prove, if she wants, that two different pseudonyms P (1)…”

“…Several such pseudonymisation approaches have been proposed in the literature (e.g. [33,24,35]). One of the most known scenarios of user-generated pseudonyms is the the case of several blockchain systems (such as the case of Bitcoin), in which the users are being identified by a meaningless identifier (i.e.…”

“…Moreover, it should be pointed out that the cryptographic strength of the above properties are strongly related to the cryptographic strength of the Merkle tree as an one-time signature scheme, which is known to be post-quantum secure under specific assumptions on the underlying hash function [9,10] (see subsection 3.3). This is an important property, taking also into account that other known techniques on deriving user-generated pseudonyms (such as the aforementioned techniques in [33,24,35])) rely on conventional public key cryptographic schemes which are not post-quantum secure.…”

Anonymization and Pseudonymization of FHIR Resources for Secondary Use of Healthcare Data

“…FHIR-DIET allows users to choose among several cryptographic schemes, all of which offer a perfectly hiding property and can be fed to the tool through the configuration files. In particular, following the guidelines [14], the choice is around literature approaches based on asymmetric cryptographic algorithms (e.g., [40]) or more recent approaches based on cryptographic accumulators for pseudonym generation [41], which inherits the security properties of a Merkle tree (also envisioned by the recent ENISA report [15]) achieving post-quantum security. The reference is the ENISA best practices [14] and techniques [15] and the ISO 25237 standard on pseudonymisation and specifically the Clause 6 [16].…”