{"id":4366,"date":"2026-11-21T15:08:12","date_gmt":"2026-11-21T15:08:12","guid":{"rendered":"https:\/\/smartcryptosecurity.com\/?p=4366"},"modified":"2026-11-21T15:08:12","modified_gmt":"2026-11-21T15:08:12","slug":"four-ciphers-in-one-full-round-attacks-on-the-reference-implementation-of-llbc","status":"publish","type":"post","link":"https:\/\/smartcryptosecurity.com\/?p=4366","title":{"rendered":"Four Ciphers in One: Full-Round Attacks on the Reference Implementation of LLBC"},"content":{"rendered":"<p>The original LLBC design was introduced in the <em>IEEE Internet of Things Journal<\/em> in 2025, praised for its low-latency capabilities tailored to resource-constrained IoT environments. To ensure reproducibility and assist developers, the designers released a reference implementation written in C and Verilog, accompanied by test vectors derived directly from this code. However, the new cryptographic analysis reveals that the released reference implementation diverges from the published peer-reviewed specification in eight distinct ways. <\/p>\n<p>The most damaging of these discrepancies lies in the round function&#8217;s rotation amounts. While the official specification calls for rotation amounts of $(3,19)$, the reference implementation uses $(32,16)$ instead. This seemingly minor coding error introduces a profound mathematical vulnerability. Kofuku and Igarashi prove that an LLBC-type cipher utilizing rotation amounts $(b_1, b_2)$ acts as a mathematical product of $g\/4$ independent ciphers when $g$, defined as the greatest common divisor of $64$, $b_1$, and $b_2$, is a multiple of $4$. Furthermore, the cipher splits entirely when $g$ becomes a multiple of $8$.<\/p>\n<p>Because of the $(32,16)$ rotation amounts implemented in the code, the reference implementation of LLBC inadvertently fractures into four completely independent 32-bit ciphers whose internal data paths never mix. Rather than operating as a unified 128-bit security primitive, the algorithm functions as four isolated lanes running in parallel.<\/p>\n<p>Exploiting this structural flaw, the researchers successfully mounted devastating attacks on all 20 rounds of the reference implementation. They devised a probability-one distinguisher requiring only two chosen plaintexts. Additionally, the team demonstrated the ability to decrypt any ciphertext using $2^32$ chosen plaintexts, and they successfully recovered the master key from just three known plaintexts using $2^80 + 3 cdot 2^16$ encryptions. This recovery cost sits below a standard exhaustive search by a massive factor of $2^48$. <\/p>\n<p>The practical implications of these theoretical breaks are striking. While the original designers estimated that the strongest possible attack against the proper specification would reach a maximum of nine rounds, the implementation flaws allowed the researchers to bypass all security mechanisms across all 20 rounds. Executing the decryption attack on a single standard personal computer required just 175.5 seconds.<\/p>\n<p>The researchers also evaluated how standard cryptographic analysis techniques perform against the flawed code. A standard automatic search successfully identified 20-round impossible differentials and zero-correlation approximations. Conversely, when examining classical differential characteristics, linear characteristics, and integral properties, the analysis only managed to distinguish 9, 10, and 8 rounds respectively within a single isolated component of the cipher. <\/p>\n<p>Furthermore, the study highlights a broader methodological risk in cryptographic engineering: the assumption inherent in usual analytical models\u2014which typically treat substitution boxes, or S-boxes, as independent units\u2014does not hold true for the flawed reference implementation of LLBC. <\/p>\n<p>In light of these findings, Kofuku and Igarashi have issued a vital recommendation for the broader cryptographic and software development communities. They advise system designers to rigorously verify released reference code against test vectors generated independently from the formal specification, rather than relying on outputs generated by the code itself. They also emphasize the necessity of thoroughly checking the diffusion properties of the final released codebase prior to formal publication.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>The original LLBC design was introduced in the IEEE Internet of Things Journal in 2025,&#8230;<\/p>\n","protected":false},"author":20,"featured_media":4365,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[676],"tags":[678,423,3430,607,514,194,3432,3433,677,3431,214,684],"class_list":["post-4366","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-cryptography-math","tag-algorithms","tag-attacks","tag-ciphers","tag-cryptography","tag-four","tag-full","tag-implementation","tag-llbc","tag-math","tag-reference","tag-research","tag-round"],"_links":{"self":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/posts\/4366","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/users\/20"}],"replies":[{"embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4366"}],"version-history":[{"count":0,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/posts\/4366\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=\/wp\/v2\/media\/4365"}],"wp:attachment":[{"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4366"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4366"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/smartcryptosecurity.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4366"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}