roadmap/2024-02-19.md at 0ce9170027d04e69340d8d1b40ce8bb5c51a7a0b

title

`network privacy and mixnet`

Based on previous investigation, started a document about integrating the population strategy from Nym into the Cryptarchia setting. In it, while following an explicit staking assumption, it is assumed that a node that wants to be a mix node must register and stake on-chain. WIP with a lot of potential changes.

Implementation WIP: defining structure for mixnode / mixclient.
Implementation: concluded that the nymtech/sphinx crate can be used for our use case. For reference, previously we used a wrapper of sphinx developed in the nymtech/nym codebase - but now, we can use the nymtech/sphinx crate as it is.
Modified libp2p to use QUIC. Replaced TCP with QUIC in the nomos-libp2p crate, which will be used by the mixnet network backend: PR.
Fixed integration tests for QUIC. With the QUIC updates, some DA integration tests were failing - now they are working properly: PR.

Created a document regarding embeddable databases with their analysis. In more detail, the document includes benchmark code, analysis of several rust-based DBs, a proposal for which DB we should use, and example code for our most important use cases.

Updated the "Does Crypsinous’ Leader Election Function lead to wealth concentration in PoS?" document with new details about stake relativization - more precisely new simulations in regards to validators of certain ranges.
Started building out a NIZK (Non-interactive Zero-Knowledge) Glossary of terms that will be used for internal education as well as writing the specification.
WIP: writing a how-to guide for reading/writing NIZK specifications. This will help us in the future when writing certain Nomos specifications.
Analysis of de-anonymization of relative stake: for adversarial inference of relative stake in the leader election process, considered statistical properties of (naive) maximum likelihood (ML) estimator of relative stake - in the naive ML approach, the relative stake of node i is obtained from the frequency of observed 1’s, representing vins in elections, and properties of lottery function. Also, analyzed statistical properties of the naive ML estimator of relative stake. The consequences of the aforementioned results for the mixnet are in progress and will be shared in the future. To see the math behind the analysis check the Overleaf document.

No tangible updates - but Cryptarchia rust implementation is in progress, more details in the coming weeks.

Initial DA API specification structure PR: defined mock zone, DA node and block producer to wrap DA spec and use it for encoding, dissemination and verification. Once the DA spec is near finalization, we will continue the mock implementation in python.
DA Specification updated per reviews and comments. Likely more changes on the way (depending on review).
Started a document on studying VeriZEXE and Taiga designs. WIP, currently includes initial notes, likely to change.

Started implementing KZG core functionality of DA - more details in the upcoming weeks.