In April 2024 the GFW began decrypting QUIC Initial packets at scale and blocking connections by domain rather than by IP address, targeting the exact transport layer that modern moving-target circumvention systems rely on for session continuity. Because blocking a registrable domain costs the censor essentially zero collateral (unlike blocking a hyperscaler IP range), the censor's optimal Stackelberg budget split allocates f* = 0.97 of its discovery budget to the domain channel once the defender runs n ≥ 2 endpoints (§8.10, Figure 7). GFWatch measurements show hundreds of thousands of domains persistently blocked.

Address blocking is a losing strategy for the censor: with n=8 endpoints and µ/λa = 1 (discovery as fast as rotation), address-denial probability falls below 10^-2, and even when the censor discovers twice as fast as the defender rotates (µ/λa = 0.5), 7 endpoints keep address denial below 10^-2 (Lemma 1, Figure 1). The collateral budget γ means the censor cannot even block all addresses it discovers, and Corollary 1 gives the explicit threshold: n ≥ ln(1/ε) / ln((λa+µ)/λa) endpoints suffice for any target ε.

§5, §8.1, §8.10, Lemma 1, Corollary 1, Figure 1 evaluation

Correlated domain burns (e.g., registrar takedowns) collapse the sustainability frontier far more than mean burn rate alone: the interval frontier β* falls from 0.65 at burst size b=1 to 0.5 at b=2, 0.3 at b=4, and at b=8 no β in the tested range is sustainable. Conversely, diversifying the same total domain economy across P=8 independent providers (where each provider's pool can be emptied by a single takedown but providers fail independently) moves the frontier past β=1, and simulated P-provider availability matches the closed form A_P = 1 - q^P geometrically (Proposition 4, Figure 5d).

§8.7, §8.8, Proposition 4, Figure 5 defense

When the domain burn rate β = λdisc/λintro exceeds 1, time-average availability is bounded by A ≤ 1/β regardless of rotation speed µ, endpoint count n, or domain buffer kmax (Theorem 2). Rotation speed µ appears only in the address factor, which redundancy already saturates, and is entirely absent from the binding domain layer — proving that 'rotate faster' is categorically the wrong design lever under a domain-filtering censor.

§6, Theorem 2 evaluation

Simulator validation shows a sharp phase transition in (α,T)-availability at the sustainability frontier β* ≤ 1. For a canonical configuration of n=8 endpoints and buffer kmax=8, the interval frontier is β* ≈ 0.65, while the time-average crossing is near β ≈ 0.8. Sweeping rotation speed µ/λa from 1 to >100 does not move the frontier — the availability contour over the (β, µ/λa) plane is entirely vertical (Figure 3). The closed-form availability law (Theorem 1) matches simulation to three significant figures across all tested β (Table 2). A rotation-only defender at beta = 1.5 stays below 0.25 availability even at the largest buffer and rotation speed tested, while a domain-aware defender at beta = 0.6 with the same n, mu and pool exceeds 0.95.

§8.2, §8.5, Figure 2, Figure 3, Table 2 evaluation

Testing the Tranco Top 100k domain list from Russian vantage points shows that all 2,644 domains blocked over TLS SNI are also blocked over QUIC SNI, and domains accessible via TLS are accessible via QUIC—indicating the TSPU maintains a unified SNI blocklist across protocols. Additionally, 31 domain patterns including discord.com exhibit 'residual clearance': QUIC connections to those domains are not censored when immediately preceded by non-censored data on the same 4-tuple; 22 domains are only blocked on port 443.

2026-heitmann-russia-s-early-introduction §3, Table 3 sni-blockinghttp3-quic-blockdpi

Reproducing Zohaib et al.'s 10-payload QUIC parser test suite against Russian TSPU devices reveals that the Russian QUIC parser differs meaningfully from the GFW: payload 8 (PING and PADDING frames inserted before the Crypto frame), which circumvents the GFW, does not bypass Russian censorship. Payload 10 (QUICv2) bypasses both censors. Payload 9 (ECH with outer SNI cloudflare-ech.com) is blocked by Russia; payload 3 (version bytes 0x00000002 with invalid MAC) is blocked by both.

2026-heitmann-russia-s-early-introduction Appendix B, Table 2 http3-quic-blockdpisni-blocking

As measured from three Russian vantage points in March–April 2026, TSPU devices decrypt QUIC Initial packets, locate the SNI extension in the embedded TLS handshake, and drop packets with forbidden domains plus all subsequent packets on the 4-tuple for 420 seconds (timer resets on each packet). Censorship is unidirectional (outgoing only), covers all destination ports up to 65535 at Saint Petersburg and Novosibirsk (Moscow exempts 22% of system ports 1–1023), and targets only QUICv1—QUICv2 passes uncensored at all vantage points.

2026-heitmann-russia-s-early-introduction §3 sni-blockinghttp3-quic-blockdpi

QUICstep successfully circumvents the GFW's QUIC SNI censorship (active since April 2024) in live testing. Using an Alibaba VM in mainland China as client and an AWS instance in North Virginia as server, a native QUIC client was blocked after several fetches of youtube.com SNI, while QUICstep consistently succeeded across 50 consecutive fetches. 7 tiktokcdn.com subdomains that were QUIC-SNI blocked were also reliably accessible via QUICstep. The approach routes only QUIC long-header (handshake) packets through a WireGuard tunnel; all subsequent short-header (data) packets travel the native path.

2026-lee-quicstep §4.3.2, §3.2.3 http3-quic-blocksni-blockingdpi

Since April 7, 2024, the GFW decrypts every QUIC client Initial packet at China's national border and blocks connections whose TLS ClientHello SNI matches a QUIC-specific blocklist. Blocking takes the form of dropping all subsequent UDP packets sharing the same (src-IP, dst-IP, dst-port) 3-tuple for 180 seconds—with no RST injection. The GFW applies a source-port heuristic: packets with src-port ≤ dst-port are not inspected, capturing >92% of real QUIC client Initials while processing only ~30% of all UDP traffic. The QUIC blocklist contains 58,207 unique FQDNs (Tranco, Oct 2024– Jan 2025), approximately 60% of the DNS blocklist in size; 33% of blocked FQDNs do not actually support QUIC, suggesting the list was derived from an existing domain-name blocklist rather than live QUIC service discovery.

2025-zohaib-quic-sni §3 / §3.1 / §3.3 / §4 http3-quic-blockdpisni-blocking

Russia's TSPU ("Средства противодействия угрозам") system is deployed inline at individual ISP edges rather than at centralized internet exchange points, producing substantial per-ISP heterogeneity: some providers apply layer-7 SNI/Host filtering while others rely primarily on IP-prefix blocklists, and QUIC/HTTP3 is blocked at several major providers. Rollout timing and enforcement depth vary measurably across autonomous systems, meaning a single "Russia passes/fails" test fixture systematically underestimates blocking coverage.

2024-xue-tspu-russia §4–§5 sni-blockingdpihttp3-quic-blockip-blocking