Master Thesis project proposal
“Designing and implementing an ECDSA-based Zero-Knowledge Credential Architecture for Yivi as EUDI Wallet”
Context and motivation
Background
Yivi is a privacy-preserving digital identity platform that has successfully launched production deployments using IRMA/Idemix protocols based on zero-knowledge proof (ZKP) schemes. With the introduction of the EU Digital Identity (EUDI) Wallet regulation (eIDAS 2.0), Yivi aims to evolve into a compliant EUDI wallet while maintaining its strong privacy guarantees and crypto-agile architecture.
The EUDI ecosystem increasingly standardises on:
-
Verifiable Credentials (e.g. W3C VC, SD-JWT-VC, ISO 18013-5 mDL/MDOC)
-
Presentation and issuance protocols such as OpenID4VCI and OpenID4VP
-
Selective disclosure and zero-knowledge techniques, analysed in detail in ETSI TR 119 476, including BBS+, CL signatures and other privacy-enhancing cryptographic mechanisms.
Yivi wants to leverage these developments while preserving its core privacy values: minimum disclosure, unlinkability, and user-controlled identity.
Strategic challenge
Today, many credentials in practice are signed using ECDSA keys (for example JWT-based credentials, SD-JWT-VC, mDL/MDOC. At the same time, privacy-preserving credential systems often rely on different cryptographic primitives (e.g. CL, BBS+ on BLS12-381).
Yivi faces a strategic challenge:
-
How to evolve towards an EUDI-compliant wallet that: Reuses existing and widely deployed ECDSA key material,
-
Supports zero-knowledge proofs and selective disclosure,
-
Remains interoperable with OpenID4VCI / OpenID4VP and standard verifiers,
-
And preserves Yivi’s strong privacy guarantees and crypto-agility.
The ECDSA-based ZKP opportunity
Recent work such as Google’s Longfellow project (“Anonymous credentials from ECDSA”) and new proposals around BBS# indicate that it is possible to:
-
Build anonymous credential schemes on top of existing ECDSA-signed credentials,
-
Provide selective disclosure and unlinkable presentations,
-
Minimise changes to issuer infrastructure,
-
And potentially integrate with standard protocols such as OpenID4VCI and OpenID4VP.
This opens the possibility for Yivi to design a next-generation ZKP layer that:
-
Uses ECDSA keys as the fundamental trust anchor,
-
Compares and possibly combines Longfellow-style constructions with BBS+/BBS#-based approaches,
-
And is grounded in the requirements and recommendations of ETSI TR 119 476.
Research objectives
Primary objective
Design and prototype an ECDSA-based zero-knowledge credential architecture for Yivi that:
-
Provides selective disclosure and unlinkable presentations based on ECDSA keys,
-
Is aligned with the cryptographic and privacy requirements from ETSI TR 119 476,
-
Supports interoperability with OpenID4VCI and OpenID4VP,
-
And can be integrated into Yivi’s roadmap towards an EUDI-compliant wallet.
Specific research questions
RQ1: Requirements analysis based on ETSI TR 119 476 How can the privacy, security and interoperability requirements from ETSI TR 119 476 for selective disclosure and ZKP-based credentials be translated into concrete requirements for a Yivi ECDSA-ZKP architecture, in particular regarding:
-
Unlinkability across presentations
-
Minimal disclosure and predicate proofs
-
Revocation and status verification
-
Crypto-agility and (future) post-quantum considerations
RQ2: ECDSA-based ZKP design options (Longfellow vs BBS#/BBS+) What are the design trade-offs between:
-
Longfellow / “Anonymous credentials from ECDSA” Using existing ECDSA-signed credentials (JWT / SD-JWT-VC / MDOC) as the base
-
Generating zero-knowledge proofs over attributes derived from these credentials
-
BBS+/BBS#-based credentials anchored in ECDSA trust Mapping Yivi (and EUDI) credential structures to BBS+/BBS# signatures
-
Exploring how ECDSA-based PKI and BBS#/BBS+-based ZKP can be combined or bridged
RQ3: Yivi architecture integration
How can an ECDSA-based ZKP scheme (Longfellow, BBS#, or a hybrid) be integrated into Yivi’s architecture while:
-
Maintaining backward compatibility with existing IRMA/Idemix credentials where needed
-
Supporting multiple credential formats (e.g. SD-JWT-VC, MDOC, IRMA) within Yivi
-
Preserving Yivi’s privacy-first design, including unlinkability and minimal disclosure
-
Allowing for crypto-agile evolution as standards mature
RQ4: Interoperability with OpenID4VCI and OpenID4VP
How can the proposed ECDSA-ZKP architecture:
-
Represent credentials and proofs in W3C VC formats (e.g. JWT/SD-JWT-VC or Data Integrity profiles)
-
Be transported using OpenID4VCI for issuance and OpenID4VP for presentations
-
Interoperate with verifiers that: Support advanced ZKP-proof types, and
-
Only support “classic” JWT/SD-JWT verification (graceful degradation / dual-path designs)
RQ5: Evaluation and recommendations
To what extent does the proposed architecture:
-
Meet the ETSI TR 119 476 criteria for privacy-preserving credentials,
-
Achieve practical performance for mobile wallets and verifiers,
-
Provide a realistic migration path for Yivi towards EUDI-compliant, ECDSA-based ZKP credentials?
What recommendations can be made to Yivi for:
-
Short-term experimentation (e.g. Longfellow-style wrapping of existing credentials), and
-
Long-term architecture choices (e.g. adoption of BBS# or hybrid designs)?
Student profile
We are looking for a motivated university-level student in Computer Science, Cyber Security or a closely related discipline. You have a strong affinity with cryptography, digital identity, and privacy-preserving technologies, and you are eager to apply academic knowledge to a real-world, high-impact use case. You work independently, think analytically, and are comfortable exploring complex technical concepts.
Thesis benefits
-
Professional supervision from specialists in cryptography, identity management, and EUDI Wallet technologies
-
Regular feedback and technical sparring sessions throughout the thesis process
-
Access to technical documentation, development environments, and research materials relevant to the assignment
-
A monthly thesis compensation of €500 (based on a 40-hour commitment; exceptions possible)
-
Flexible working arrangements, including hybrid work options
-
Opportunities to publish or present your research within the organization
-
Real-world impact: your work may directly contribute to the integration of Yivi as an EUDI Wallet
References
Academic
-
Anonymous credentials from ECDSA https://eprint.iacr.org/2024/2010
-
Privacy-Preserving Credentials: Camenisch et al https://eprint.iacr.org/2014/468.pdf
-
ETSI TR 119 476 - Electronic Signatures and Trust Infrastructures (ESI); Analysis of selective disclosure and zero-knowledge proofs applied to Electronic Attestation of Attributes https://www.etsi.org/deliver/etsi_tr/119400_119499/119476/01.02.01_60/tr_ 119476v010201p.pdf
BBS# and eIDAS 2.0 https://csrc.nist.gov/csrc/media/presentations/2024/wpec2024-3b3/images-media/wpec2024-3b3-slides-antoine-jacques- -BBS-sharp-eIDAS2.pdf
Other
-
What is Yivi https://docs.yivi.app/what-is-yivi
-
IRMAGO https://github.com/privacybydesign/irmago
-
EUDI Wallet ARF: EU Commission - Regulatory framework https://eudi.dev/2.5.0/architecture-and-reference-framework-main/
Contact
Primary contact person
Dibran Mulder, CTO Caesar Groep & Yivi
+31 (0)6 39 30 61 18
[email protected]
Address:
Janssoniuslaan 80
3528 AJ Utrecht
Websites:
https://yivi.app
https://caesar.nl
