In mathematics, the correlation immunity of a Boolean function is a measure of the degree to which its outputs are uncorrelated with some subset of its inputs. Specifically, a Boolean function is said to be correlation-immune of order m if every subset of m or fewer variables in ${\displaystyle x_1, x_2, \ldots, x_n}$ is statistically independent of the value of ${\displaystyle f(x_1,x_2,\ldots,x_n)}$.

## Definition

A function ${\displaystyle f:\mathbb{F}_2^n\rightarrow\mathbb{F}_2}$ is ${\displaystyle k}$-th order correlation immune if for any independent ${\displaystyle n}$ binary random variables ${\displaystyle X_0\ldots X_{n-1}}$, the random variable ${\displaystyle Z=f(X_0,\ldots,X_{n-1})}$ is independent from any random vector ${\displaystyle (X_{i_1}\ldots X_{i_k})}$ with ${\displaystyle 0\leq i_1<\ldots.

## Results in cryptography

When used in a stream cipher as a combining function for linear feedback shift registers, a Boolean function with low-order correlation-immunity is more susceptible to a correlation attack than a function with correlation immunity of high order.

Siegenthaler showed that the correlation immunity m of a Boolean function of algebraic degree d of n variables satisfies m + d ≤ n; for a given set of input variables, this means that a high algebraic degree will restrict the maximum possible correlation immunity. Furthermore, if the function is balanced then m + d ≤ n − 1.[1]

## References

1. Template:Cite journal

Template:Crypto-stub

de:Correlation immunity