P2P Privacy
Research
- Major effort was devoted to advancing the Ordering Experiments descriptions to cover more complex scenarios, to unblock the progress of simulations. The work was successful and all scenarios have been described, discussed, and implemented. The whole section contains six generalized experiments that cover three different topological settings with two mix node behaviors. The set of parameters was adjusted to be able to finish on time based on a set of benchmarks that we ran. More on that here.
- The Experiments Tracking document was updated to include new experiments. A decision was made to run experiments from the most comprehensive and computationally heavy (6th) down to the lightest (1st). This way we can consider making an analysis quicker and wait for supportive results.
- Implemented and ran the Ordering Measurements. Implementation can be found at this branch. A PR hasn’t been opened yet since some parts need to be cleaned up, and we wanted to quickly implement the new experiments. However, we took the time to check that the experiments work as intended. Also, ran the benchmark to estimate the duration of experiments.
- We are currently running all the experiments (in order from heaviest to lightest). All of them are expected to be done by 6th September.
- Analysis of latency in the Nomos mix network: considered a node sending a message to all N−1 nodes of the random network. A message traveling from a node to its neighbor is delayed by the amount of time spent in the out-queue plus the connection link latency. Defined the distance between nodes i and j as the minimum latency of a message sent from node i to node j. For node 1, the maximum overall distance to N−1 nodes is the latency of broadcast. Using this definition, considered the latency of broadcast on the network with the topology of a complete graph. The latter is a special case of a random regular graph. Developed code to compute the latency of broadcast for networks with the topology of random regular graphs. This work is summarized in this document. Analysis of a message sent from a node in a random regular graph is currently in progress.
Development
- No updates this week.
Data Availability
Research
- No updates this week.
Development
- PR #691 (Node) - created the aggregated protocols behavior swarm and implemented the DA network backend with the aforementioned libp2p swarm.
- PR #693 (Node) - updated DaBlob struct and related code to include column index.
- PR #694 (Node) - added an ability to store and load DaBlobs with different column indexes in DA Storage (used by validator, indexer, will be required by the sampling service).
- PR #695 (Node) - implemented the DA network backend with the above libp2p swarm.
- PR #698 (Node) - implemented subnetworks membership algorithms: - v1: filling with available nodes.
- PR #701 (Node) - executor libp2p backend implementation: swarm implementation for Nomos Cli app using libp2p executor behavior- PR #702 (Node) - implemented subnetworks membership algorithms: - v2: filling with available nodes with extra replication on the original data columns subnetworks.
PPoS/Consensus
Research
- Continued development of the code to study the selfish behavior of stakers (within the wealth concentration context).
- Integrated the statistical analysis of the “follow the protocol” study (within the wealth concentration context).
Blockers: None
Development
- No updates this week.
Coordination Layer
Research
- Concluded the VM analysis: after reviewing the candidate zkVMs, we decided that continuing with Risc0 is the most suitable option at this stage for our Native Zone needs. SP1, Jolt, Cairo-VM, and Nexus will be among the VMs we continue to monitor, but Risc0 appears to be the best choice for now.
- PR #31 and PR #33 (PoCs) - performed optimization of native zones: use basepoint as the blinding point and provide state roots instead of whole state.
- PR #32 (PoCs) - built isolated proof scenarios to aid benchmarking work.
- PR #34 (PoCs) - aggregated input/output proofs into a single partial transaction proof.
- PR #35 (PoCs) - moved the blinding to the partial transaction level instead of blinding each note value individually.
- PR #36 (PoCs) - switched from Pederson balance commitments to hash balance commitments.
- PR #37 (PoCs) - modelled zone transaction log with MMR.
Development
- No updates this week.
Testnet + Insights
Development
- No updates this week.
User Tools
Research
- No updates this week.
Development
- No updates this week.
Miscellaneous
- No updates this week.