In a single-round censorship game the only Nash equilibrium that keeps the channel open requires the circumvention traffic proportion (CTP) satisfy CTP ≤ F, where F = (βant+βbnt)/(αact+αbct+βant+βbnt). In repeated indefinite games a stable equilibrium exists at CTP = Z = (1−p)·CTPmax, where p is the per-round continuation probability, allowing a non-zero proportion of circumvention traffic to flow indefinitely without triggering shutdown.

Adding a DPI apparatus with true positive rate TPR and false positive rate FPR creates three ordered thresholds Fam ≤ Fab ≤ Fmb governing censor strategy: allow all traffic (CTP ≤ Fam), deploy the apparatus (Fam < CTP ≤ Fmb), or block all traffic (CTP > Fmb). The apparatus does not qualitatively change the Nash equilibrium structure; it only shrinks the CTP range the circumventor can sustain, with the ordering Fmb ≥ Fab ≥ Fam holding whenever TPR ≥ FPR.

§4.3 detection

A censor can mount a zero-collateral-damage flooding attack by injecting fake CRS-protocol-conformant traffic into open channels, inflating the apparent CTP and evicting real circumvention traffic to throttled or sacrificial protocols. If injection is costless the censor can drive real circumvention throughput to zero while keeping all channels nominally open; the attack is equally effective against both throttling and dumping CTP control strategies.

§6.1 detection

The optimal multi-protocol CRS traffic allocation distributes circumvention traffic across n cover protocols proportionally to each protocol's non-circumvention traffic volume (CTPi = Li · CTP/(1−CTP)), keeping every individual protocol below the blocking threshold. This makes individual protocol channels independently optimizable, with the sole selection criterion being maximizing cover traffic volume L rather than any other protocol property.

§5.1 defense

Throttling—capping total CRS traffic at Fab and withholding surplus—strictly dominates dumping surplus traffic onto a sacrificial protocol that will subsequently be blocked. Table 2 shows that at CTP = Fab·1.05 the circumventor's relative utility drops to 0.88 of the Fab baseline when dumping, while throttling preserves all open protocols; under a censor flooding attack dumping additionally loses protocol n entirely, making throttling the dominant strategy in both attack and no-attack conditions.

§5.1.3, Table 2 defense

Stage 1 of the detection pipeline uses a lightweight heuristic: restrict analysis to IP addresses in "VPS-dense ASNs," which censors already target for resource-intensive inspection of fully-encrypted traffic. This pre-filter dramatically reduces the search space before applying the more expensive dual-role behavioral analysis. The evaluation was conducted without Stages 1 and 3 due to dataset limitations, meaning the reported 23% recall and 0.18% FPR are conservative lower bounds on the full pipeline's performance.

2026-almutairi-server §2.1, §2.2, §3.4 fully-encrypted-detectdpiml-classifier

Adversarial pre-padding — prepending stochastic byte noise to packets — degrades ET-BERT encrypted traffic classification accuracy from >99% to 25.68%, exposing a structural vulnerability in all payload-byte-dependent detection systems. White-box adversarial attacks (Ayaka AH-MSI) additionally achieve evasion rates exceeding 99.5% against standard continuous-time sequence models via Manifold Shattering, where adversaries align malicious temporal distributions with benign baselines.

2026-ferrel-aegis-adversarial-entropy-guided §II, §VI-A dpiml-classifier

Explicitly disentangling packet headers (structured, low-entropy) from encrypted payloads (high-entropy, stochastic) into separate MoE branches yields consistent gains across six datasets: 86.85% F1 on 120-class TLS 1.3 traffic (CSTNET-TLS), 97.88% F1 on USTC-TFC2016 malware/benign flows, and 92.65% F1 on imbalanced IoT traffic (CIC-IoT2022), demonstrating that headers and payloads carry fundamentally different and independently exploitable discriminative signals.

2026-he-trafficmoe-heterogeneity-aware-mixture §III-A, §IV-B, Tables II–III ml-classifierdpifully-encrypted-detect

Striding with factor 4 (early downsampling) produces the largest single-factor degradation in the ablation study: average macro-F1 drops from 0.9909 to 0.9772 and cross-dataset variance increases from 4.77×10⁻⁵ to 4.51×10⁻⁴, with worst-case dataset performance falling to MIN 0.9524. Fine-grained byte order and short-range structure — protocol headers, payload signatures, repeated byte motifs — carry essential discriminative signal that stride-based aggregation destroys.

2026-kulatilleke-mambanetburst-direct-byte-level §V-B, Table IV ml-classifierdpi

A burst of just 5 packets truncated to 320 bytes each (1600 bytes total) suffices for macro-F1 ≥0.9824 across all six benchmarks; the classification token reads from the final recurrent state after a 4-layer Mamba-2 stack processing this fixed-length prefix, with no additional flow-level or session-level context required.

2026-kulatilleke-mambanetburst-direct-byte-level §III-A, §V-A ml-classifierdpi

Classification from the first 5 packets × 320 bytes (1600-byte burst) achieves near-perfect accuracy across Tor (F1=0.9990), VPN (F1=0.9871), malware (F1=0.9954), and IoT attack traffic (F1=0.9966), with IP addresses masked and only header and initial payload retained. The earliest portion of each packet provides sufficient discriminative information for a classification decision made within the first kilobyte of a flow.

2026-kulatilleke-mambanetburst-direct-byte-level §III-A, §V-A, Table II–III ml-classifierdpi