Experiment Protocol

Experiment Protocol

Benchmark scope

The repository includes BBS and AnonCreds summary benchmark files over attribute counts:

4, 8, 16, 32, 64

The measured protocol represented by the summary files uses 5 warmup runs and 50 measured runs.

Metrics

The result tables track:

• prove or present latency;
• verify latency;
• proof bytes;
• VP bytes;
• peak RSS memory;
• scaling ratio;
• per-attribute growth;
• selector decision outcome;
• rejection reasons.

Reproducibility path

Run:

python experiments/run_all.py

This regenerates derived tables, selector outputs, ablation results, resource-sensitivity outputs, figures, and synthetic raw traces.

Synthetic trace warning

The synthetic raw-run file is generated from summary statistics. It is useful for testing and plotting but must not be described as original raw cryptographic trace data.