Cryptarchia §
Research §
Development §
Mixnet (Network Privacy) §
Research §
- Defined metrics to be collected for Mixnet (for other components as well, but mainly for Mixnet). This list should be sufficient for now.
- Investigated which empirical analysis had been done by Loopix paper and another paper, in order to design Nomos empirical analysis soon. When compared to our previous work, it isn’t so different. The main goal would be measuring the probability of linking an output message from a source message correctly. Still remains to determine if this should be for executable spec, or for a real implementation.
- Planning the first collaboration with DST. This would focus on the Mixnet properties, but only simple metrics as the first step. Before starting the collaboration, we need to check if this list is really reasonable to be offloaded to DST.
- In consideration on how we can avoid node distinction and registration, and while reviewing the Hopr design an idea arose of changing the layered mix topology (Nym) to a circular one which enables users to extend their number of hops freely, and using gossip pub-sub protocol as an underlying messaging to achieve mixing over P2P. Therefore, we are able to have at the same time privacy (and scalability) optimal layered mix (with flexibility of adjusting the level of anonymity - path length) that does not require registering node long-lasting public identifiers (IPs) and works on top of a P2P network. All this led to a sketch of a new type of a mix, the “Whirl” Mix.
- New Network Privacy proposal has been made, based on the updated and clear vision/roadmap of Nomos.
- Optimization of mixnets: considered a scenario when all N nodes, where at most M nodes are adversarial, are used to construct the mixnet with L layers. Proved that in this case the fraction of compromised paths 𝛼α is at most (𝑀/𝐿)𝐿(M/L)L which was also verified by simulations. The histogram of 𝛼α obtained in simulations only has a long left tail. Properties of the above mixnet are very different from properties of mixnets of size n sampled from N nodes. Here the fraction of compromised paths is always bounded by 1. This suggests that for small (or moderate) N using all nodes could be more desirable. The summary of this work is provided in Notion while the details of the analysis can be found in Overleaf.
Development §
Data Availability §
Research §
Development §
- Merged KZG+RS core (implementation of the KZG and RS core methods): PR.
- Implemented and merged DA V1 encoder: PR.
- Implemented and merged DA V1 verifier: PR.
- DA mempool specific functionality for cert to vid conversion: network payload and mempool item in DA mempool now can differ if payload can be converted into the mempool item, PR.
- DA Indexer (WIP): implementation of metadata indexer had to be postponed as it depended on the certificate, vid and metadata definitions added in mempool split PR and da-protocol-v1 branch.
Coordination Layer §
Research §
- Studied on the proof aggregation methods and also on the Jolt protocol (previously reviewed) for the survey. Added the updates (WIP - the aim is to complete it by the end of this week). A brief explanation about Jolt: It is designed to reduce prover costs by using lookup arguments. The reason for reviewing it is that its implementation was released in the past weeks. This version uses the Hyrax polynomial commitment, which results in higher verifier complexity (which is not ideal for Nomos). However, they mentioned that future versions would support different commitment schemes. Our priority is minimizing verifier costs, but if Jolt can offer good results for both prover and verifier, it might be worth considering. Release post of Jolt on X.
Development §
- Split generic mempool into CL and DA mempools - this was needed as making it completely generic would have been impossible at some point. They are now free to deviate in both typing and behavior without affecting the other. The refactor was done so they can share as many available resources as possible: PR.
Testnet §
Development §
Miscellaneous §
- WIP: Nomos Roadmap and Milestone Execution Plan.
- Wealth Concentration in PoS, part 1 - published and posted on X.
- Introduction to Nomos Architecture - in final stages, will be shared by Wednesday or Thursday.