Censors optimize for utility under asymmetric misclassification costs rather than raw accuracy: false positives (blocking legitimate traffic) carry economic and political costs that make censors conservative about deploying classifiers with high false-positive rates. Multi-flow stateful classifiers — such as the obfs4 Elligator probabilistic distinguisher, which requires correlating observations across multiple connections — are operationally more expensive than single-packet or connection-initiation classifiers, which the author suggests explains why probabilistic multi-flow distinguishers have not been exploited in practice even when theoretically available.

Despite fully encrypted protocols existing since obfs2 in 2012, the first documented evidence of the GFW passively detecting them purely by randomness appeared only in 2021 — approximately a decade later — and was limited to certain foreign IP address ranges and a subsampled fraction of traffic. Meanwhile, the GFW had been discovering obfs2/obfs3 servers via active probing as early as 2013, indicating censors found active-probing-based address discovery cheaper and more reliable than passive statistical classifiers for this protocol family.

§5 evaluation

Three independent implementation flaws in obfs4proxy's Elligator encoding made obfs4 public-key representatives passively distinguishable from uniform random bytes: (1) non-canonical square roots allowed a square-then-root test matching 100% of obfs4 outputs but only ~50% of random strings; (2) bit 255 was always zero; (3) only large prime-order subgroup points were encoded. A classifier exploiting these achieves 100% sensitivity (obfs4 never falsely marked as random) at less-than-100% specificity. All three were fixed in obfs4proxy-0.0.12 (December 2021) and 0.0.14 (September 2022).

§3 detection

Shadowsocks 'stream cipher' methods lacked integrity protection on ciphertexts, enabling a decryption oracle: an attacker who can guess as few as 4 bytes of plaintext prefix (5 bytes without controlling a /24) can replay a recorded session with a modified 7-byte target header, causing the server to send the decryption of the entire recorded stream to an attacker-controlled host. This provides an efficient active test for identifying Shadowsocks servers; once identified, a censor can block by IP address.

§2 detection

VMess's encrypted command block used a non-keyed hash over variable-length fields in a MAC-then-encrypt construction where the receiver cannot locate the hash without first parsing the protected data, enabling an active distinguishing attack: by replaying an authentic request 16 times with the padding-length field P set to 0000–1111, an attacker observes that a VMess server reads exactly P+N+4 bytes before disconnecting, with max and min byte counts differing by exactly 15 with every intermediate value present. V2Ray mitigated this in v4.23.4 by disconnecting after a timeout rather than after receiving a full command block.

§4 detection

Internal Geedge documents confirm active contracts to deploy GFW-derived censorship and surveillance infrastructure in Myanmar, Pakistan, Ethiopia, Kazakhstan, and at least one additional unidentified country under the Belt and Road framework, in addition to domestic deployments in Xinjiang, Jiangsu, and Fujian. The exported product (the Tiangou Secure Gateway / TSG line) is not a stripped-down export variant — leaked TSG documentation shows DPI, active-probing, ML classifiers, and granular per-region traffic control rules that mirror the domestic GFW capability set.

2025-geedge-mesa-leak §5, §6 dpiactive-probingml-classifiersni-blockingdns-poisoningfully-encrypted-detecttraffic-shape

InterSecLab's 76-page analysis of the Geedge/MESA leak (based on nine months of indexing and translating >100,000 documents) characterizes the Tiangou Secure Gateway (TSG) product line as a commercially deployable detection stack that combines deep packet inspection, real-time mobile subscriber monitoring, active probing, ML-based traffic classifiers, and granular per-region rule sets. TSG is not a research prototype — leaked documentation includes deployment timelines and client government interactions for Kazakhstan, Ethiopia, Pakistan, Myanmar, and one unnamed country, with censorship rules explicitly tailored to each region.

2025-interseclab-internet-coup §3–§5 (InterSecLab report) dpiactive-probingml-classifiersni-blockingip-blockingtraffic-shapefully-encrypted-detect

Censorship classifiers and traffic analysis attacks consistently exploit the initial seconds of a proxy connection, where packet-size, inter-arrival-time, and burst features are maximally discriminative. Cited work demonstrates that website fingerprinting classifiers trained solely on the first few seconds of Tor traffic achieve high accuracy, and real-world GFW detection of fully-encrypted protocols also targets early-connection bytes.

2025-pereira-extended §1 traffic-shapewebsite-fingerprintml-classifierdpifully-encrypted-detect

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

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

Balboa's synchronous leaf-content replacement adds non-negligible timing differences that allow censors to identify its activity with up to ~90% accuracy over different network conditions. The timing anomaly arises because Balboa performs data substitution directly at each data exchange, delaying the server's response while covert data is prepared.

2026-kamali-huma §I (Introduction), §II-A traffic-shapeml-classifier