Skip to content

Security: everanium/itb

SECURITY.md

ITB Security Reference

Security notice. ITB is an experimental symmetric cipher construction without prior peer review, independent cryptanalysis, or formal certification. The construction's security properties have not been verified by independent cryptographers or mathematicians.

PRF-grade hash functions are required. No warranty is provided.

ftotal = fpixel ∘ ftraj ∘ fcobs ∘ fslot ∘ frank   ⟹   ftotal−1 ∘ ftotal = id
Geometric Bijection

No bespoke cryptography. ITB composes established, standardized primitives rather than introducing new cryptographic designs. Security properties and regulatory status are inherited from the underlying primitives; see README.md for jurisdictional certification details.

Companion document to SCIENCE.md. This file is the audit-oriented operational reference — composition modes matrix, cross-scheme comparisons, deployment threat model, design-rationale tables, compliance framing, and the explicit non-closed threats list. The formal construction and proofs live in SCIENCE.md, PROOFS.md, and the machine-checked Lean 4 suite in lean/README.md; the accessible-explanation narrative lives in ITB.md. Sections cross-referenced below point to the authoritative source for each topic.

Security Policy

Supported versions. Active development tracks main. Security fixes land on main first; older tags receive backports only when the maintainer judges the fix worth backporting (rare given the pre-release status).

Version Supported
main (HEAD) ✓
latest tagged releases ✓

Reporting a vulnerability. Private reports preferred via GitHub Security Advisories or by email at andrew@encloud.blue. Discord channel for coordination: discord.gg/wRYF8shHpd.

Triage and response. Best-effort triage within 14 days. Pre-release status means no SLA or commercial support is implied. Responses are individually authored, not template-driven.

Coordinated disclosure. Findings against the construction — cryptanalytic attacks, theoretical weaknesses, implementational flaws — are most welcome and prepared for academic publication. A 90-day coordinated disclosure window is the default; faster public disclosure may be appropriate when a fix lands quickly or the finding is already public via parallel discovery. The ITB construction was published to invite scrutiny, not to claim closed-form security — cryptanalytic findings under the author's own attribution are explicitly encouraged.

Out of scope. Hardware-level attacks (DPA/SPA, Spectre, Meltdown, Rowhammer, cache timing, fault injection), supply-chain attacks against dependencies, attacks against the host OS / language runtime / hardware platform itself. See the Security notice above and SCIENCE.md § 7.

1. ITB Composition Modes

Property Core ITB (No MAC) MAC + Silent Drop [1] MAC + Reveal
Integrity ✗ ✓ ✓
Deniability ✓ Full (structural) ✓ Full ✓ Full (full-capacity MAC)
CCA oracle No oracle exists No oracle (silent) Noise position only (noiseSeed)
noiseSeed config ✓ Barrier intact ✓ Barrier intact ✗ Leaked via CCA
dataSeed config (per-region) ✓ Barrier intact ✓ Barrier intact ✓ Independent (zero CCA leak)
lockSeed config ✓ Barrier intact ✓ Barrier intact ✓ Independent (zero CCA leak)
startSeed config (per-region) ✓ Barrier intact ✓ Barrier intact ✓ Independent (zero CCA leak)
Data rotation + XOR ✓ ✓ ✓ (rotation barrier)
Desynchronized Interlocked Barrier — Rank Barrier & Layer II slot desynchronization (0x03/0x06, non-disableable) [2] ✓ Intact ✓ Intact ✓ Intact (lockSeed independent of noiseSeed)
Desynchronized Interlocked Barrier — Pixel Barrier & Layer I sub-region trajectory (0x05, non-disableable) [2] ✓ Intact ✓ Intact Partial — noise position leaked; rotation + XOR + DRBG residue intact via dataSeed, Layer I trajectory intact via startSeed
Brute-force impact of leak — — noiseSeed eliminated [3]: P × 2^(2·keyBits) → P × 2^keyBits

[1] MAC + Silent Drop assumes the attacker is unaware of MAC presence. If the attacker knows a MAC is inside (insider knowledge), the encrypted MAC tag serves as a local verification oracle during brute-force: candidate keys are verified by decrypting and checking MAC(payload) without recipient interaction. Search cost remains P × 2^(2·keyBits) (identical to Core ITB — no CCA leak, noiseSeed not revealed; both seeds must be searched jointly). Without insider knowledge: no verification predicate → plausible deniability. Grover bound: √P × 2^keyBits.

[2] Software-level property under the random-container model; no guarantees against hardware-level attacks (see Security notice above and SCIENCE.md § 7). Information-theoretic absorption under passive observation applies to the Pixel Barrier (P(v | h) = 1/256, Proof 1); the Rank Barrier's per-chunk mask permutation is PRF-conditional and keyed by lockSeed (Proof 11). In the Desynchronized Interlocked Barrier, startSeed_i (for region r = i − 1 ∈ {0, 1, 2}) keys Layer I sub-region trajectory desynchronization under domain tag 0x05 and lockSeed keys Layer II cross-lane chunk-slot desynchronization under domain tag 0x06, both folding each 128-bit output pair (hLo, hHi) via the bijection fullWidthFold(hLo, hHi) (fullWidthFold_bijection, fullWidthFold_preserves_prf_cardinality in lean/ITB/Desync.lean; SCIENCE.md § 1.2.1) and both isolated from noiseSeed (zero CCA leak). Both barriers run together as an indivisible composition (SCIENCE.md § 1.2).

[3] CCA eliminates noise bits (12.5 %), but CSPRNG-seeded DRBG fill bytes remain encrypted in data bit positions by dataSeed, indistinguishable from plaintext (Proof 10).

For the underlying PRF requirements, 8-seed isolation proof, and multi-factor KPA argument, see SCIENCE.md § 1.1 ChainHash, § 2.3 8-Seed Isolation, § 2.6 Multi-Factor Full KPA Resistance. For cross-cipher key-size and primitive-requirement comparisons and the AEAD comparison matrix, see SCIENCE.md § 4 Comparison with Existing Ciphers. For the byte-splitting property analysis, see SCIENCE.md § 2.10 Byte-Splitting Property.

2. CCA Oracle Leak Comparison

Scheme CCA Leak (MAC result revealed)
AES-CBC + MAC-then-Encrypt [1] Padding oracle → full plaintext (POODLE, Lucky13)
AES-CTR + MAC-then-Encrypt Bit-flip oracle → data structure
ITB + MAC-Inside (full capacity) Noise position only (3 bits / pixel, no data); DRBG fill persists in data positions (Proof 10)
AES-GCM (AEAD) None (MAC rejects before decryption)
ChaCha20-Poly1305 (AEAD) None (MAC rejects before decryption)

[1] Deprecated construction (TLS 1.0 / 1.1, RFC 8996). Included as an architectural reference baseline only.

For the CCA leak upper bound proof and ITB's noise-position-only bound, see SCIENCE.md § 2.8 CCA Leak Upper Bound and Proof 6.

3. Attack Resistance Summary

The Desynchronized Interlocked Barrier (SCIENCE.md § 1.2) is the composite pair of the per-chunk Rank Barrier permutation (default 48-bit drawn from a ≈ 2^70.20 mask space, or 168-bit drawn from a ≈ 2^258.60 mask space under Interlock = "168", paired with Layer II cross-lane chunk-slot desynchronization Σ_0, Σ_1, Σ_2 under domain tag 0x06 with β_0 ≠ β_1, multiplying single-chunk cross-lane slot matching by M^3 to |Ω_chunk| · M^3 ≈ 2^96.91–2^101.53 in 48-bit mode and ≈ 2^279.89–2^284.51 in 168-bit mode, or ≈ 2^128.91–2^133.53 and ≈ 2^311.89–2^316.51 with the 4-byte interlock-nonce fold v at the 1024-bit ITB Mode 2 / Mode 1 floors) and the per-pixel Pixel Barrier channel-encoding stage (paired with Layer I dynamic sub-region trajectory desynchronization Π_r under domain tag 0x05 with K_r ∈ [6, 10] in Mode 2 and K_r ∈ [8, 16] in Mode 1, yielding a 3-region trajectory space of ≈ 2^166.98 on the 1024-bit ITB Mode 2 floor and ≈ 2^224.04 on the Mode 1 floor at K_r = 8, leaving > 2^145 and > 2^198 trajectories for fixed initial entry pixel Π_r(0)). It applies uniformly across every column in the table below; the barrier's contribution to the KPA / Crib KPA / CPA columns is architectural and PRF-conditional, and stacks on top of the pixel-level obstacles listed per cell.

Attack Core ITB MAC + Silent Drop [5] MAC + Reveal MAC + Reveal + KPA
Ciphertext-only (COA) ✓ IT barrier [1] ✓ IT barrier [1] ✓ IT barrier [1] ✓ IT barrier [1]
Crib KPA ✓ Anchor destroyed [2] ✓ Anchor destroyed [2] ✓ Anchor destroyed [2] ✓ Anchor destroyed [2]
Full KPA ✓ 4-factor under PRF [2] ✓ 4-factor under PRF [2] ✓ 4-factor under PRF (noisePos leaked; 7^P intact) [2] ✓ 4-factor under PRF (noisePos leaked; 7^P intact) [2,3]
Partial KPA ✓ 5-factor under PRF (byte-splitting auxiliary) [2] ✓ 5-factor under PRF [2] ✓ 5-factor under PRF [2] ✓ 5-factor under PRF [2,3]
Chosen-plaintext (CPA) ✓ Fresh per-message maps & barrier draws ✓ Fresh per-message maps & barrier draws ✓ Fresh per-message maps & barrier draws ✓ Fresh per-message maps & barrier draws
Chosen-ciphertext (CCA) ✓ No oracle ✓ No oracle noiseSeed leaked; dataSeed / lockSeed / startSeed safe [4] noiseSeed leaked; dataSeed / lockSeed / startSeed safe [4]
Brute-force (classical) P × 2^(2·keyBits) [5] P × 2^(2·keyBits) [5] P × 2^keyBits [6] P × 2^keyBits [6]
Brute-force (Grover) √P × 2^keyBits [5] √P × 2^keyBits [5] √P × 2^(keyBits/2) [6] √P × 2^(keyBits/2) [6]
Map guessing 2^(62P) 2^(62P) 2^(59P) 2^(59P)
Nonce reuse Lab-only; Rank Barrier scrambles XOR [7] Lab-only; Rank Barrier scrambles XOR [7] Lab-only; Rank Barrier scrambles XOR [7] Lab-only; Rank Barrier scrambles XOR [7]
Bit-flipping Undetected Undetected (silent drop) Detected (MAC) Detected (MAC)
Padding oracle N/A (no padding) N/A (no padding) N/A (no padding) N/A (no padding)
Quantum structural (Simon, BHT) Conjectured mitigated Conjectured mitigated Conjectured mitigated Conjectured mitigated

[1] Software-level property under the random-container model; no guarantees against hardware-level attacks (see Security notice above and SCIENCE.md § 7).

[2] Full KPA resistance is 4-factor under the PRF assumption (Theorem 4a): (1) PRF non-invertibility (candidate verification requires key), (2) startPixel isolation and Layer I sub-region trajectory desynchronization (untransmitted; under domain tag 0x05, K_r ∈ [6, 10] in Mode 2 and K_r ∈ [8, 16] in Mode 1 yield a 3-region trajectory space of (2^7 · 8! · 18^8)^3 ≈ 2^166.98 at K_r = 8 on the Mode 2 floor — > 2^145 for fixed initial entry pixel Π_r(0) across the 147^3 ≈ 2^21.6 entry-pixel combinations, and > 2^104 at K_r = 6 — and (4^7 · 8! · 51^8)^3 ≈ 2^224.04 at K_r = 8 on the Mode 1 floor — > 2^198 for fixed initial entry pixel Π_r(0) across the 408^2 · 409 ≈ 2^26.02 entry-pixel combinations, and up to > 2^445 at K_r = 16), (3) per-pixel encoding ambiguity (56^P without CCA; under CCA Reveal, noise positions leak, leaving 7^P rotation ambiguity), and (4) Rank Barrier per-chunk mask permutation and Layer II cross-lane slot desynchronization (≈ 2^70.20 balanced partitions for default 48-bit or ≈ 2^258.60 for 168-bit, Theorem 11; anti-collapse preserved by gcd(A, B) = 66,861 for 48-bit or gcd(A₁₆₈, B₁₆₈) = 10,078,335,949,946,663,736 for 168-bit two-step reduction, Theorem 12; under Layer II domain tag 0x06, guaranteed distinct lane shifts β_0 ≠ β_1 multiply single-chunk slot matching by M^3 to |Ω_chunk| · M^3 ≈ 2^96.91–2^101.53 in "48" and ≈ 2^279.89–2^284.51 in "168" at the Mode 2 / Mode 1 floors, or ≈ 2^128.91–2^133.53 and ≈ 2^311.89–2^316.51 with the 4-byte fold v). Byte-splitting (gcd(7, 8) = 1) acts as an auxiliary 5th factor specifically under Partial KPA (each channel mixes bits from 2 adjacent bytes; unknown adjacent bytes prevent channel candidate formulation). Under Full KPA, brute-force targets seeds directly so byte-splitting adds no additional factor. In Crib KPA, the Rank Barrier removes fixed byte-position anchors. See SCIENCE.md § 1.2.1, § 2.6, § 2.10, § 2.16, and Proof 4a.

[3] Per-bit XOR hides XOR masks under passive observation. With an invertible primitive, naive pixel-level inversion would attempt seed recovery, but the Desynchronized Interlocked Barrier's Rank Barrier permutation and Layer I/II desynchronization (0x05 / 0x06, folded via fullWidthFold) foreclose the SAT anchor at instance formulation (see REDTEAM.md § FNV-1a lo-lane SAT).

[4] CCA reveals noise bits (12.5 % of container), but CSPRNG-seeded DRBG fill bytes encrypted by dataSeed persist in data bit positions, indistinguishable from plaintext. The information-theoretic barrier is reduced, not eliminated (Proof 10).

[5] Core ITB and MAC + Silent Drop (no oracle): the attacker must jointly search noiseSeed and dataSeed — without dataSeed, noiseSeed output is indistinguishable from random, so independent attack on noiseSeed is impossible without knowledge of dataSeed (Proof 3). Joint search space: 2^(2·keyBits). When enumerating full seeds, startSeed contributes the initial entry-pixel factor P (Π_r(0), enumerated per region as [0, region_width)) alongside its 0x05-keyed Layer I sub-region trajectory schedule (note [2]). Total: P × 2^(2·keyBits) classical, √P × 2^keyBits Grover. At 1024-bit keys under the default per-region container (Mode 1, P = 1225): classical ~2^2058, Grover ~2^1029 (under per-container Mode 2, P = 441: classical ~2^2057, Grover ~2^1028).

[6] MAC + Reveal: CCA reveals noisePos but neither startPixel nor the Layer I sub-region trajectory (both determined by independent startSeed + nonce under tags 0x02 and 0x05, not transmitted; initial entry pixel enumerated per region as [0, region_width)). Total: P × 2^keyBits classical, √P × 2^(keyBits/2) Grover. At 1024-bit keys under the default per-region container (Mode 1, P = 1225 = 3 × per-region floor 365, square-rounded to 35 × 35 with DefaultBarrierFill = 1): classical ~2^1034, Grover ~2^517 (under per-container Mode 2, P = 441: classical ~2^1033, Grover ~2^516). See [2,3] above and SCIENCE.md § 2.6.

[7] The Rank Barrier removes the demask entry point on colliding messages: the per-chunk PRF-keyed 48-bit (or 168-bit) mask triple (0x03) and Layer II cross-lane chunk-slot permutations Σ_0, Σ_1, Σ_2 (0x06) are applied before COBS framing, so two-time-pad XOR yields slot-permuted, lane-compressed bits p₀(P₁) ‖ p₁(P₁) ‖ p₂(P₁) ⊕ p₀(P₂) ‖ p₁(P₂) ‖ p₂(P₂) interpretable only with lockSeed and PRF-opaque under the PRF assumption. Empirically, plaintext recovery is null under every attacker-realistic dual-slot, main-only, or interlock-only collision probe at tested sample sizes (REDTEAM.md § Nonce reuse). Under single-slot partial collisions (main-only or interlock-only), the uncollided fresh axis closes confidentiality a fortiori. See SCIENCE.md § 1.2, § 2.14 Nonce Uniqueness, § 2.15, and PROOFS.md § Nonce Uniqueness.

Nonce reuse is local. The shipped wire carries two independent nonces: public main_nonce on the wire header and interlock_nonce inside the container lane buffers. Birthday collision on either single slot reaches ~50 % after 2^(N/2) messages (~2^256 for default 512-bit, ~2^64 for 128-bit); simultaneous dual-slot collision requires the product probability ~2^N messages (~2^512 for default 512-bit). Within one message (64 MB cap, about 2^23.4 chunks of 48 bits, or about 2^21.61 chunks of 168 bits) a repeated per-chunk mask is expected with probability ~2^-24 (or ~2^-216 under 168-bit interlock). Both are drawn per encrypt from crypto/rand and enforced pairwise-distinct at the API, so caller-side reuse is not reachable through the shipped surface. Even under a lab-forced collision, seeds remain secret (PRF non-invertibility blocks ChainHash inversion), so future messages with fresh nonces are unaffected — no key rotation required. A single nonce collision provides too few observations for Simon, BHT, or quantum structural algebraic attacks. Unlike AES-GCM where nonce reuse leaks the GHASH key H and enables forgery until key rotation (a global catastrophe affecting all subsequent messages), ITB nonce collision is strictly local — ITB is nonce-misuse-resistant under the PRF assumption. For usage-precondition framing, see ITB.md § 9 Nonce Reuse.

4. Information-Theoretic Barrier Metrics

Container Bit Accounting (per pixel)

Metric Bits Percentage
Total container bits 64 100 %
Data bits 56 87.5 %
Noise bits 8 12.5 %

Configuration Bit Accounting (per pixel: noiseSeed + dataSeed)

Source Config bits CCA leak Protected
noiseSeed (noise position) 3 3 (100 % of noiseSeed) [1] 0
dataSeed (rotation + XOR) 59 0 (independent seed) 59 (100 %)
Total 62 3 (4.8 %) 59 (95.2 %)

[1] CCA reveals noise bit positions, leaving DRBG fill encrypted in data positions (Proof 10; see § 1 [3]).

Barrier Strength (1024-bit key)

Every entry point applies the CCA-resistant MinPixels floor uniformly, so the auth-side envelope covers Core ITB, MAC + Silent Drop, and MAC + Reveal — one row per metric, no plain-vs-auth split.

Metric Mode 2: Per-Container Mode 1: Per-Region (Default) Theoretical Floor (Single Region)
Container dimensions 21 × 21 (P = 441) 35 × 35 (P = 1225) 20 × 20 (P = 400)
MinPixels enforcement 365 joint → 441 (DefaultBarrierFill = 1) 3 × 365 = 1095 → 1225 (DefaultBarrierFill = 1) 365 → 400
Noise barrier (2^(8P)) 2^3528 2^9800 2^3200
Encoding ambiguity 56^P (No CCA, Theorem 9) 2^2561 2^7114 2^2323
Encoding ambiguity 7^P (Under CCA, Theorem 9) 2^1238 2^3439 2^1123
Guaranteed DRBG residue (Theorem 10) ≥ 287 bytes (s = 20) / 532 bytes ≥ 483 bytes (s = 34) / 910 bytes ≥ 273 bytes (s = 19) / 245 bytes
Landauer bound (blind enumeration) ~2^306 ~2^306 ~2^306
Blind-enumeration exponent vs Landauer 11.5× (3528 / 306) 32× (9800 / 306) 10.5× (3200 / 306)
Config map space (2^(62P)) 2^27342 2^75950 2^24800
Key space 2^1024 2^1024 2^1024
Mask-space cardinality per chunk (Theorem 11) ≈ 2^70.20 ≈ 2^70.20 ≈ 2^70.20
PRF-preimage count per mask triple ≈ 2^57.80 ≈ 2^57.80 ≈ 2^57.80
gcd(A, B) anti-collapse factor (Theorem 12) 66,861 (full Cartesian reached) 66,861 (full Cartesian reached) 66,861 (full Cartesian reached)
Mask-space cardinality per chunk, 168-bit (Theorem 11) ≈ 2^258.60 ≈ 2^258.60 ≈ 2^258.60
PRF-preimage count per mask triple, 168-bit (512-bit PRF) ≈ 2^253.40 ≈ 2^253.40 ≈ 2^253.40
gcd(A₁₆₈, B₁₆₈) anti-collapse factor, 168-bit (Theorem 12) 10,078,335,949,946,663,736 (full Cartesian reached) 10,078,335,949,946,663,736 (full Cartesian reached) 10,078,335,949,946,663,736 (full Cartesian reached)
Layer I 3-region trajectory (tag 0x05) ≈ 2^166.98 at K_r = 8 (≈ 2^126.00 at K_r = 6) ≈ 2^224.04 at K_r = 8 (> 2^445 at K_r = 16) (3-region mode)
Layer II cross-lane slot space |Ω| · M^3 (tag 0x06) ≈ 2^96.91 ("48") / ≈ 2^279.89 ("168"); ≈ 2^128.91 / ≈ 2^311.89 with 4 B v ≈ 2^101.53 ("48") / ≈ 2^284.51 ("168"); ≈ 2^133.53 / ≈ 2^316.51 with 4 B v (3-lane mode)

Note on Mode 2. Under the per-container mode (Mode 2), MinPixels is evaluated jointly across the whole container rather than per region, and every barrier and ambiguity figure in its column exceeds the 2^1024 key space.

Note on Mode 3 (Gauss-Posit). Under the inscribed Gauss-circle mode (Mode 3, Mode = 3), the 1024-bit container floor is ⌊(6 × 365) / π_posit(es=2)⌋ + 1 = 698 → 28 × 28 (P = 784, P_0 = P_1 = 261, P_2 = 262) with DefaultBarrierFill = 1 (6324 B wire floor at 128-bit nonce vs 9852 B in Mode 1, or 6372 B vs 9900 B at 512-bit nonce, saving 3528 B / 35.8%): noise barrier 2^(8P) = 2^6272 (20.5× the 2^306 Landauer exponent); encoding ambiguity 56^784 ≈ 2^4553 (No CCA) and 7^784 ≈ 2^2201 (Under CCA); guaranteed DRBG residue ≥ 385 bytes (s = 27, 602 bytes at the 698-pixel floor); config map space 2^(62P) = 2^48608; classical brute-force P × 2^2048 ≈ 2^2057.6 (Core ITB) / P × 2^1024 ≈ 2^1033.6 (MAC + Reveal); Layer I (tag 0x05, posit64(es) + quire512 chord schedule, K_r ∈ [6, 10]) 3-region trajectory space 2^141–2^233 (≈ 2^141.15–2^233.39, ≈ 2^187.16 at K_r = 8, leaving > 2^162 at K_r = 8 for fixed initial entry pixel Π_r(0) across 261² · 262 = 17,847,702 ≈ 2^24.09 entry-pixel combinations); and Layer II (tag 0x06) cross-lane slot space |Ω| · M^3 at M = 880 ("48") and M = 251 ("168"): ≈ 2^99.54 (or ≈ 2^131.54 with the 4-byte interlock-nonce fold v) in "48" and ≈ 2^282.52 (or ≈ 2^314.52 with fold) in "168".

The Landauer row bounds the cost of blind enumeration of the noise-barrier space; it does not bound structural attacks that do not enumerate.

The noise-barrier headline is stated at the unified MinPixels floor. For the noise-barrier bound derivation, see Proof 5; for the per-chunk mask-space derivation, see Proof 11; for the gcd anti-collapse trap derivation, see Proof 12.

Practical Value of 4.8 % CCA Leak

Information Gained by attacker?
Plaintext bits Zero
XOR mask bits Zero
Start pixel & Layer I sub-region trajectory (0x02 / 0x05) Unknown — startSeed_i independent of noiseSeed
Interlocked Barrier — Rank Barrier & Layer II cross-lane slot desynchronization (0x03 / 0x04 / 0x06) Zero — lockSeed independent of noiseSeed; MAC verified strictly before COBS decode and Layer II invLaneDesync (desync_preserves_verify_before_cobs); per-chunk mask triple and cross-lane slot permutations Σ_0, Σ_1, Σ_2 stay PRF-opaque (drawn from ≈ 2^70.20 balanced partitions per chunk for 48-bit, or ≈ 2^258.60 for 168-bit, multiplied by M^3 Layer II slot triples)
Key-space reduction noiseSeed eliminated: P × 2^(2·keyBits) → P × 2^keyBits
Brute-force speedup Search space halved in exponent (two seeds → one seed); the barrier's per-chunk mask enumeration (≈ 2^70.20 masks per chunk for 48-bit, ≈ 2^258.60 for 168-bit), Layer II cross-lane slot desynchronization (× M^3 per single-chunk slot match), and Layer I sub-region trajectory space stack on top
Grover reduction √P × 2^keyBits → √P × 2^(keyBits/2) (noiseSeed eliminated from search); the barrier's per-chunk mask enumeration and Layer I/II desynchronization are not amenable to Grover (no observable anchor to search against)
DRBG residue after CCA Persists: fill bytes in data positions encrypted by dataSeed (Proof 10)

5. Noise-Density Optimality (Why 8/1)

Format Data / px Noise / px Overhead CCA Config Leak Barrier (1024-bit, unified P)
8/1 (ITB) 56 8 1.14× 4.8 % 2^3200
6/2 48 16 1.33× 8.9 % 2^6400
5/3 40 24 1.60× 12.2 % 2^11616
4/4 32 32 2.00× 17.1 % 2^16928

8/1 is Pareto-optimal among the analyzed noise-density configurations. The Barrier column is illustrative: 2^(noise/px × P) at each format's theoretical floor P = ⌈√MinPixels(format)⌉², MinPixels(format) = ⌈1024 / log₂(data bits per channel)⌉ (treated as total-container floor for the illustrative comparison across formats). The shipped ITB construction over-provisions this in Mode 1 by applying MinPixels per region × 3, giving P = 1225 for 1024-bit keys with barrier 2^9800 — see §4. All noise-barrier exponents place blind enumeration above the Landauer bound; that scopes to enumeration cost, not to structural attack resistance.

6. MAC Placement Design Space

MAC Placement Covers Noise Deniability CCA Leak Circular Dependency
Inside (full capacity) No ✓ Preserved Noise position only None
Inside (plaintext only) No ✓ Preserved Noise pos + spatial None
Outside (header) Yes ✗ Broken None Verification oracle
Inside (full container) N/A N/A N/A Tag invalidates itself

Implemented: Inside (full capacity) — every shipped MAC Authenticated Low-Level entry point (EncryptAuthenticated3x{128,256,512}Cfg) and every triple.Pipeline profile that enables MAC. Other placements are theoretical alternatives, not implemented. For the composition analysis, see SCIENCE.md § 2.13 MAC-Inside-Encrypt Composition.

7. Known Theoretical Threats

Threat Exploit requires Practical risk Mitigation
Interlock combinadic unrank index-select timing (DPA/SPA class) Oscilloscope on CPU die, > 10 GHz, lab access Same class as DPA on any cipher Constant-time VPERMT2Q / VPERMD select over precomputed C(p, k) table, no software side-channel
rotateBits7 shift timing (DPA/SPA class) Oscilloscope on CPU die, > 10 GHz, lab access Same class as DPA on any cipher Register-only, no software side-channel
Container size metadata Network observation Metadata only Inherent to all ciphers, no crypto advantage
Non-DRBG container Deployer misconfiguration Degrades barrier internal/drbg (CSPRNG-seeded, or crypto/rand direct under csprng) mandatory, container source outside it unsupported
COBS decode truncation Wrong seed / tampered data None Core ITB: returns raw decoded bytes (plausible deniability, no oracle); MAC Authenticated: MAC rejects before COBS decode
Bit-flip false null (DoS) Data bit modification None (with MAC) MAC verified before null search; noise flips do not affect decrypted data
CGO AVX2 side-channel Co-located attacker None (see below) All AVX2 ops constant-time; identical to Pure Go
Spectre v1 / v2 / v4, Downfall, etc. Secret-dependent memory access gadget No known gadget in ITB data path Register-only ops; no table[secret_index]
MDS, Zenbleed (stale data) CPU buffer residue Seeds may remain in buffers Not ITB-specific; identical for all ciphers
Rowhammer, RAMBleed DRAM bit flips / reads Memory corruption / leakage Not ITB-specific; ECC memory recommended
Heap memory exposure Memory dump, debugger, Meltdown Seeds, cached hash keys in heap Not ITB-specific; explicit Pipeline.Close() and Cipherset.Close(), secureWipe on pooled/intermediate buffers and SoEM22 states, wipeInterlockWorker48 / wipeInterlockWorker168 on per-worker Rank Barrier scratch/masks, and KDF/MAC/DRBG/keystream/wrapper state zeroization

For detailed per-CVE analysis of 20+ hardware attacks (Spectre variants, Downfall, Hertzbleed, MDS, Zenbleed, Rowhammer), see HWTHREATS.md.

CGO Backend Side-Channel Equivalence

The optional C pixel processing backend (CGO_ENABLED=1, GCC -O3, SIMD tiers selected at runtime) was verified for side-channel equivalence with the Pure Go backend:

Concern Status Detail
AVX2 / AVX-512 / GFNI / VBMI / AES-NI / BMI2 instruction timing Constant-time Fixed latency across supporting Intel / AMD microarchitectures (BMI2 on AMD Zen 1 / Zen 2: microcoded) for all five pixel tiers (Tiers A, A′, B, B′, C) and hash / interlock SIMD kernels. See HWTHREATS.md § Category 5 for the complete per-instruction inventory and microarchitecture latency table
dataHash % 7 Constant-time GCC optimizes to imulq multiply-by-reciprocal, no div instruction
% totalPixels (pixel wrap) Variable-time idivl Not secret: totalPixels = W × H from public header
Container access pattern Same as Pure Go container[pixelOffset] — contiguous per-sub-region cache pattern under Layer I (subregionAddr, effectiveStartPixel) identical to Pure Go
Hash array access Sequential No data-dependent indexing
Spectre Not applicable No secret-dependent array indexing
L1 micro-batching Improves cache resistance 8 KB batches harder to observe via Flush+Reload than full arrays

The analysis applies equally to ARM64 NEON auto-vectorization: veor, vand, vorr, vshl, vshr are constant-time on ARM. ARM sdiv (for % totalPixels) is variable-time but operates on public data only. ARM has no frequency throttling from NEON (unlike Intel AVX-512).

Both backends produce identical ciphertext. Switching between CGO_ENABLED=0 (Pure Go pixel kernel; Go-assembly hash ASM stays engaged) and CGO_ENABLED=1 (C + SIMD pixel kernel) does not change the security model on any platform.

8. Hash Function Compliance

Every shipped registry primitive is a construction that satisfies the requirements catalogued in SCIENCE.md § 1.1 ChainHash and § 2.6 Multi-Factor Full KPA Resistance. Tests and benchmarks cover every registry entry across the 128 / 256 / 512-bit widths shipped by the ChainHash surface; the concrete registry list is authoritative in hashes/registry.go. For per-primitive technical notes across the shipped registry constructions, see the hashes/CONSTRUCTIONS.md reference.

For key-size and primitive-requirement comparisons vs AES / ChaCha20 / Threefish, see SCIENCE.md § 4 Comparison with Existing Ciphers. For the effective key-size table by hash width (128 / 256 / 512), see SCIENCE.md § 6 Implementation and ITB.md § 4 8-Seed Isolation.

9. Threat models NOT closed by the barrier

The classes below define the boundary of what the Interlocked Barrier addresses — either sitting outside its threat model or holding conditionally on a usage or primitive precondition:

Closure of the KPA / CPA families is conditional on a secure PRF and fresh nonces; total inversion or a reused nonce is outside what the barrier closes.

  • Total or systematic PRF inversion. The barrier is explicitly PRF-conditional. Total inversion of the configured primitive lets the attacker resolve the per-chunk masks and Layer II cross-lane slot shifts, the startPixels and Layer I sub-region trajectories, and the seed components algorithmically, collapsing every architectural obstacle. Systematic partial PRF inversion is raised in cost but not eliminated. Outside the barrier's power to close.
  • Nonce reuse. Closure of the CPA / KPA families is conditional on fresh per-message nonces on both the main nonce and interlock nonce slots (§ 3 [7]). The shipped API draws both internally per call from crypto/rand (so simultaneous collision requires a CSPRNG hardware fault), which is an API-discipline property rather than a construction-level guarantee. Empirically, lab-forced dual-slot collision recovers zero plaintext bytes at the tested sample sizes (see REDTEAM.md § Nonce reuse).
  • Upstream key-management / related-seed supply. The 8-seed API draws independent CSPRNG components and rejects both pointer and byte-level Components collisions, so related-seed differentials do not arise through the shipped surface. A defective upstream KDF that supplies correlated seeds is outside the construction's control. Threat-model boundary, not a barrier property.
  • Physical side channels (timing, power, EM, cache, speculative execution). Out of scope by construction — leakage is a property of the execution environment, not the algorithm. The barrier's own kernels are written to a constant-time discipline (mask-merge only, no secret-dependent branches or memory accesses; the deterministic reduction avoids rejection sampling precisely to stay branch-free), but a deployment-level side-channel evaluation is per-CPU, per-compiler, and orthogonal to the algorithm-level claims. See HWTHREATS.md for the per-instruction inventory.
  • Key-material compromise. If the seeds or the derived state leak through any channel, confidentiality is lost by definition. The barrier assumes secret seeds.
  • Implementation defects in the shipped Go code. The architectural argument assumes the code correctly implements the construction. Bugs in the unrank kernels, the reduction, the interleave, or the seed plumbing would be defects the architecture cannot self-correct. The test suite demands byte-level asm ↔ Go parity, directed reduction-boundary vectors at the exact gcd / preimage-cut points, and mask disjointness / popcount assertions rather than roundtrip tests alone.
  • Fault injection. Active perturbation of the running computation is a hardware-execution threat outside the algorithm's model.
  • CCA via a deployment decryption oracle. Core ITB and MAC + Silent Drop expose no oracle; MAC + Reveal is an optional deployment mode whose CCA behaviour is per-deployment and bounded by Proof 6 (noise position, 3 bits per pixel). Whether an oracle exists at all is a protocol choice, not a barrier property.
  • Quantum key recovery. The seed-space search under Grover is a hypothetical upper-bound argument, not something the barrier addresses directly; the oracle-degradation framing applies (under MAC + Silent Drop there is no verification predicate for Grover to run against). Post-key quantum recovery is outside the barrier's threat model. For the full quantum resistance and combinatorial cryptanalysis, see SCIENCE.md § 3.3 Quantum Resistance and POSTQUANTUM.md.
  • Undiscovered cryptanalysis. No finite argument, architectural or empirical, establishes resistance to techniques not yet known. The construction has had no external cryptanalysis or formal certification. For the scope and maturity disclaimer, see SCIENCE.md § 7 Scope and Maturity Disclaimer.

References

  • Nadhem J. Al Fardan and Kenneth G. Paterson, "Lucky Thirteen: Breaking the TLS and DTLS Record Protocols", 2013 IEEE Symposium on Security and Privacy (S&P), pp. 526–540, 2013, DOI: 10.1109/SP.2013.42.
  • Bodo Möller, Thai Duong, and Krzysztof Kotowicz, "This POODLE Bites: Exploiting The SSL 3.0 Fallback", Google Security Advisory, September 2014 (PDF).
  • Kathleen Moriarty and Stephen Farrell, "Deprecating TLS 1.0 and TLS 1.1", IETF RFC 8996, March 2021, DOI: 10.17487/RFC8996 (RFC Editor).
  • Rolf Landauer, "Irreversibility and Heat Generation in the Computing Process", IBM Journal of Research and Development, Vol. 5, No. 3, pp. 183–191, 1961, DOI: 10.1147/rd.53.0183.
  • Ronald A. Fisher and Frank Yates, "Statistical Tables for Biological, Agricultural and Medical Research", Oliver and Boyd, London, 1938 (6th ed., 1963); and Richard Durstenfeld, "Algorithm 235: Random permutation", Communications of the ACM, Vol. 7, No. 7, p. 420, July 1964, DOI: 10.1145/364520.364540.
  • Stuart Cheshire and Mary Baker, "Consistent Overhead Byte Stuffing", IEEE/ACM Transactions on Networking, Vol. 7, No. 2, pp. 159–172, 1999, DOI: 10.1109/90.769765 (PDF).
  • Carl Friedrich Gauss, "De nexu inter multitudinem classium, in quas formae binariae secundi gradus distribuuntur, earumque determinantem", 1834, in Carl Friedrich Gauss Werke, Band II: Höhere Arithmetik, Königliche Gesellschaft der Wissenschaften zu Göttingen, pp. 269–291, 1863 (Cambridge Library Collection – Mathematics, Cambridge University Press, 2011), DOI: 10.1017/CBO9781139058230.012.
  • John L. Gustafson and Isaac T. Yonemoto, "Beating Floating Point at its Own Game: Posit Arithmetic", Supercomputing Frontiers and Innovations, Vol. 4, No. 2, pp. 71–86, June 2017, DOI: 10.14529/jsfi170206 (PDF); and Posit Working Group, "Standard for Posit Arithmetic (2022)", March 2022 (PDF).
  • Lov K. Grover, "A fast quantum mechanical algorithm for database search", Proc. 28th Annual ACM Symposium on Theory of Computing (STOC '96), pp. 212–219, 1996, DOI: 10.1145/237814.237866 (arXiv:quant-ph/9605043).
  • Daniel R. Simon, "On the Power of Quantum Computation", SIAM Journal on Computing, Vol. 26, No. 5, pp. 1474–1483, 1997, DOI: 10.1137/S0097539796298637.
  • Gilles Brassard, Peter Høyer, and Alain Tapp, "Quantum Cryptanalysis of Hash and Claw-Free Functions", LATIN '98: Theoretical Informatics, LNCS Vol. 1380, pp. 163–169, Springer, 1998, DOI: 10.1007/BFb0054319 (arXiv:quant-ph/9705002).
  • Paul Kocher et al., "Spectre Attacks: Exploiting Speculative Execution", 2019 IEEE Symposium on Security and Privacy (S&P), pp. 1–19, 2019, DOI: 10.1109/SP.2019.00002 (spectreattack.com, arXiv:1801.01203).
  • Moritz Lipp et al., "Meltdown: Reading Kernel Memory from User Space", 27th USENIX Security Symposium (USENIX Security 18), pp. 973–990, 2018 (Communications of the ACM, Vol. 63, No. 6, pp. 46–56, 2020), DOI: 10.1145/3357033 (meltdownattack.com, arXiv:1801.01207).
  • Daniel Moghimi, "Downfall: Exploiting Speculative Data Gathering", 32nd USENIX Security Symposium (USENIX Security 23), pp. 7179–7193, 2023 (downfall.page, USENIX).
  • Yingchen Wang, Riccardo Paccagnella, Elizabeth Tang He, Hovav Shacham, Christopher W. Fletcher, and David Kohlbrenner, "Hertzbleed: Turning Power Side-Channel Attacks Into Remote Timing Attacks on x86", 31st USENIX Security Symposium (USENIX Security 22), pp. 679–697, 2022 (hertzbleed.com).
  • Tavis Ormandy, "Zenbleed (CVE-2023-20593)", Google Information Security, July 2023 (Advisory).
  • Yoongu Kim et al., "Flipping Bits in Memory Without Accessing Them: An Experimental Study of DRAM Disturbance Errors", 41st ACM/IEEE International Symposium on Computer Architecture (ISCA 2014), pp. 361–372, 2014, DOI: 10.1109/ISCA.2014.6853210 (Project Zero).
  • Leonardo de Moura and Sebastian Ullrich, "The Lean 4 Theorem Prover and Programming Language", Automated Deduction – CADE 28, LNCS Vol. 12699, pp. 625–635, Springer, 2021, DOI: 10.1007/978-3-030-79876-5_37 (PDF).

There aren't any published security advisories