Section 11 - BBS vs PS - Asymptotic Analysis

[sampolgar]

[sampolgar]

Summary of the meeting yesterday (please correct me if I misunderstood or left anything out!)

Problem: private credential issuing and usage for an identity holding hierarchical credentials (master and context credentials). The setting is where a user has a high security credential with strong accountability (e.g. can easily be revoked) and context credentials linked to it that don't need the same level of security. E.g. passport connected to driver's license. Or master workplace credential connected to multiple service credentials.

In order of priority

Write comparison table of Anonymous Credential systems including threshold issuance, sybil resistance, revocation, pseudonyms/multiple credentials
Analysis of managing multiple credentials for issuing and usage: our analysis will compare/benchmark the methods to link credentials (public like CanDID, equality of commitments, 2 methods to do prf)
Analysis of BBS+ and PS (algebraic/practical) with issuing/usage scenario
Current focus is on the linking primitive in the context of CanDID. Next is implementing the entire system
Background thoughts

understand the setting i.e. are we in the CanDID setting or do we have a different enhancement to the layer 0 (not urgent)
if we want this to be separate primitive, think about the clean abstraction syntax
Current progress

CanDID Improvement - Accountable Privacy https://www.overleaf.com/3272253211fpvhkhpqvxcg#f87160
This was the previous document, I had trouble trying to implement and understand everything together and created a new project to manage to narrow the focus on the hierarchy structure

Hierarchical Credentials
https://www.overleaf.com/6125329759jtmtvntcydvq#7fb197
I am working through 2. from above (analysis of the different ways to link credentials) in this paper and will update the group. (The current writeup is under heading Sam and Tian 15/11)

Implementation (https://github.com/sampolgar/anonymous-credentials/) this is the project I'll be working on, nothing much to see here atm.