Voiceover's DCGAN, trained on ~400 hours of two-person telephone conversations, generates conversation timing templates that constrain when the tunnel transmits audio. This reduces ML classifier performance from auROC 0.981/aucPR 0.959 (unshaped baseline) to auROC 0.682/aucPR 0.482, and the improvement holds at 500-packet windows (auROC 0.68/aucPR 0.50), suggesting robustness to memory-limited adversaries.

Variable bitrate encoding (e.g., the OPUS codec's 6–510 kbps range) in VoIP protocols leaks content properties through packet timing, enabling ML classifiers to distinguish protocol tunnels from real conversations. An audio tunnel without timing shaping was identifiable with auROC 0.981 and aucPR 0.959 by an AutoGluon-Tabular classifier examining 1000-packet flow windows.

§2, §4.2 detection

Protocol mimicry that replicates only statistical or syntactic traffic properties is insufficient for unobservability: Houmansadr et al. (2013) showed SkypeMorph was trivially detectable by the absence of Skype control channels, missing login-server communication, and failure to replicate implementation-specific bugs present in real Skype—demonstrating that full behavioral replication, not just traffic shaping, is required to withstand scrutiny.

§2 detection

Skype for Web normalizes packet sizes such that Voiceover transmissions and genuine audio conversations produce nearly identical packet size CDFs across Ubuntu 18.04 and Windows 10, across all tested modulation parameters (carrier frequency, sampling frequency, baud rate, frame length). This makes the Skype-based tunnel inherently immune to packet-size fingerprinting without requiring explicit size shaping.

§4.4 evaluation

Voiceover achieves 31.16 bytes/s goodput with default parameters—roughly half the 62.32 bytes/s of the unshaped baseline—because GAN-imposed silence periods reduce transmission time. Skype's OPUS codec bounds the theoretical ceiling at 750–63,750 bytes/s, so all multimedia tunnels over this path are constrained to low-bandwidth use cases; the authors explicitly position Voiceover as an out-of-band channel for sharing secret keys rather than a general-purpose data path.

§4.5, Table 3 evaluation

NATA (Non-invasive Active Traffic-correlation Analysis) injects low-frequency bandwidth waveforms (sinusoidal, square-wave, triangular) into Tor TCP connections at an upstream gateway without endpoint compromise, payload decryption, or Tor-browser modification. BM-Net, a selective state-space classifier trained on the exit-side observations, achieves a 99.65% binary detection F1 score distinguishing watermarked from natural traffic on a 20,000-trace real-world dataset.

2026-fan-activeflowmark-assessing-tor §IV, §VI.D, Table III flow-correlationtraffic-shapeml-classifier

BM-Net achieves a 99.65% binary detection F1 score for distinguishing bandwidth-watermarked Tor flows from natural traffic, outperforming all evaluated baselines (next best: TikTok at 75.96% F1). The accuracy gap stems from active perturbation imposing a deterministic low-frequency throughput constraint rather than relying on subtle natural metadata, making the detection task fundamentally easier than passive website fingerprinting.

2026-fan-activeflowmark-assessing-tor §VI-D, Table III–IV flow-correlationtraffic-shapeml-classifier

BM-Net achieves a 99.65% binary detection F1 score distinguishing watermarked from natural Tor flows, and a 97.5% macro-F1 score for fine-grained modulation classification across sinusoidal, square-wave, and triangular patterns. The fine-grained test set contains 201 held-out samples collected from ten clients across five geographic regions (Europe, North America, Australia, Southeast Asia, East Asia), with training traces including traffic collected under WTF-PAD and Walkie-Talkie defenses.

2026-fan-activeflowmark-assessing-tor §VI-D, Table III, Table V flow-correlationml-classifiertraffic-shape

BM-Net achieves a 97.5% macro-F1 score for fine-grained classification of three modulation geometries (sinusoidal, square-wave, triangular) from noisy exit-side Tor observations using only 201 labeled test samples collected across cross-continental Tor paths. Residual errors concentrate between natural traffic and square-wave modulation, as abrupt low-rate transitions are partially smoothed by Tor multiplexing and network jitter.

2026-fan-activeflowmark-assessing-tor §VI-D, Table V flow-correlationml-classifiertraffic-shape

Fine-grained modulation classification (natural vs. sinusoidal vs. square-wave vs. triangular) achieves 97.5% macro-F1 on a 201-sample held-out test set. Square-wave waveforms are the hardest class (F1 = 95.7%), while sinusoidal and triangular each reach 99.0% F1, because abrupt square-wave transitions are partially smoothed by Tor multiplexing and network dynamics.

2026-fan-activeflowmark-assessing-tor §VI.D.2, Table V flow-correlationtraffic-shapeml-classifier

The host-profiling censor (passive traffic analysis: count connections per server, block those exceeding a threshold τ within a window w) blocks essentially all circumvention user traffic within 30 time steps for all classifier qualities tested (ρ_TP ∈ {0.9, 0.95, 0.99}), while causing far less collateral damage than zig-zag (never exceeding ~30% innocent server blocking). Snowflake resists this attack well: with w=3, τ=3, over 94.48% of users receive a proxy within 2 steps even with worst-classifier rules, and final unblocked server rates are 91.24–99.04%. The host profiling approach is feasible for passive censors who cannot enumerate the distribution channel.

2026-fares-game §5.2, Table 2, Table 3 traffic-shapeml-classifier