Master Thesis Project | ECDSA Zero-Knowledge Credentials in Yivi’s EUDI Wallet

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