- The simulation runs multiple Nomos Blend nodes using netrunner.
- Each node implements the Nomos Blend Tier 1~3 protocols.
- Persistent Transmission
- Message Blending
- Cover Traffic
- Also, each node generates data messages probabilistically (similar as the consensus protocol).
- The simulation prints logs that can be used to analyze the following properties.
- Latency
- Bandwidth
- Anonymity
The simulation can be configured by config/blendnet.json.
We recommend setting the step time equal to the minimum latency in the network settings, in order to speed up the simulation. For example, if all regions are used, and if the minimum latency is 40ms (asia:asia), we recommend setting the step time to 40ms.
The simulation runs until the wards.sum
number of messages (data or cover) arrive in
their last blend node and finally fully unwrapped.
The simulation constructs a network with the node_count
number of nodes.
Each node establish connections with the connected_peers_count
number of nodes randomly selected.
"persistent_transmission": {
"max_emission_frequency": 1.0,
"drop_message_probability": 0.0
},
We recommend setting max_emission_frequency
to 1 message per second,
which is the same as the expected maximum message frequency configured in the consensus protocol
that will use the Nomos Blend protocol.
To disable drop messages, set drop_message_probability
to 0.
"number_of_mix_layers": 2,
"max_delay_seconds": 10
The number_of_mix_layers
is a parameter for the Cryptographic Processor. And, the max_delay_seconds
is for Temporal Processor, which is the same as
"epoch_duration": "432000s",
"slot_duration": "20s",
"slots_per_epoch": 21600,
"number_of_hops": 2,
These parameters are defined in the specification.
In short, each node selects the slots_per_epoch / node_count * (1 / number_of_hops)
number of slots
at the beginning of each epoch. At every selected slot, the node generates a cover message.
cargo build --release
../target/release/blendnet-sims --input-settings ./config/blendnet.json
The simulation prints a bunch of logs that can be used for analysis. We recommend piping logs to a file.
To analysis logs for latency and anonymity, see the README in the scripts.
The bandwidth consumed by each node is printed as a log at the end of the simulation.