CCLab is an adversarial testing framework showing learning-based congestion controllers are generally more robust than traditional human-designed ones under feature- and environment-level attacks, with adversarial traces usable to train improved controllers.
Learning-based vs human- derived congestion control: An in-depth experimental study,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
CCLab: Adversarial Testing of Learning- and Non-Learning-Based Congestion Controllers
CCLab is an adversarial testing framework showing learning-based congestion controllers are generally more robust than traditional human-designed ones under feature- and environment-level attacks, with adversarial traces usable to train improved controllers.