Boolean Functions for Cryptography
Boolean functions — built from nothing but zeros and ones — sit at the heart of block ciphers, stream ciphers and hash functions. The strength of these systems depends on choosing functions with the right cryptographic properties. This site collects results, data and reference material from ongoing research into the design and analysis of cryptographic Boolean functions.
What's here
A Boolean function of $n$ inputs is a map $f : \{0,1\}^n \to \{0,1\}$ — its truth table is a string of $2^n$ bits. Despite this simple definition, finding functions that resist all known attacks remains an open problem. Researchers measure a function against criteria such as nonlinearity, correlation immunity, algebraic degree and autocorrelation, and group functions into affine equivalence classes that share these properties.
Explore
- The Basics: a worked introduction, from truth table to Walsh spectrum, nonlinearity and affine equivalence.
- Equivalence Classes — complete affine equivalence class tables for $n = 3, 4, 5$, and a full summary of the 150,357 classes for $n = 6$.
- Class Graph Explorer: an interactive map of how the classes connect for $n = 3, 4, 5$, with a hill-climbing animation.
- Function Analyser: paste any truth table (up to 10 inputs) to get its nonlinearity, degree, Walsh spectrum, algebraic immunity and class.
- Class Sizes: the self-mapping constant, which counts the functions in every class, with exact counts by nonlinearity for $n = 6$.
- Efficient Implementation — the word-packing and transform techniques that make exhaustive Boolean function research feasible.
- Constructions and Search: a bent function construction, the Boolean terrain, dynamic hill climbing, and the history of heuristic search at QUT.
- S-boxes and Linear Redundancy: why all 255 component functions of the AES S-box are affine equivalent (check it in your browser), and a survey of cipher S-boxes.
- Publications — papers on cryptographic Boolean function analysis and linear redundancy in the AES S-box.
- Learning — references and resources for getting into the field.
About
This work began as PhD research at the Queensland University of Technology (Information Security Research Centre) and continues today. The original Boolean Planet site accompanied that research and was widely used as a reference for the $n = 6$ classification — data that researchers have continued to draw on, including NIST's 2018 study of the multiplicative complexity of 6-variable Boolean functions. This is its modern home.
About me
I completed my PhD on Boolean functions for cryptography at the Queensland University of Technology in 2003, and since then my work has spanned:
- Data science: I'm currently a Senior Data Scientist at Optimum, analysing Optimum Network performance.
- Formal verification of blockchain systems at Consensys, where I co-authored papers on deductive verification of smart contracts in Dafny (arXiv 2208.02920), a formal and executable semantics of the Ethereum Virtual Machine in Dafny (arXiv 2303.00152), and formal verification of the Ethereum 2.0 Beacon Chain.
- Financial econometrics at QUT's School of Economics and Finance: forecasting volatility and electricity price spikes.
- Teaching at university and secondary level.
My interest in Boolean functions never stopped. I've kept coming back to them over the years, and I have unpublished work that I'm keen to finalise and share. This site is a first step.
To get in touch, please message me on LinkedIn.
My research interests: cryptographic Boolean functions and their classification; affine equivalence and invariants; cryptanalysis.
By Dr Joanne Fuller · ORCID 0000-0002-5221-3505 · Google Scholar · LinkedIn · PhD thesis (PDF)