Function

This package is used to evaluate fitness values for backdoors.

Guess-and-determine

Implementation of the fitness function from (…) to find weakening sets of variables (aka backdoors). The fitness value is calculated on a random sample of tasks that is generated in Sampling module. For each task, random substitutions are selected for variables from the evaluated backdoor. If necessary, additional correct values of variables are also generated (for example, when processing cryptographic functions).

The behavior of this implementation is controlled by the following parameters:

• budget – An instance of the Budget module.

• measure – An instance of the Measure module.

from function.impl import GuessAndDetermine

function = GuessAndDetermine(
    budget: Budget
    measure: Measure
)

Inverse Backdoor Sets

Implementation of the fitness function from [AAAI2018] designed to search inverse backdoor sets in cryptographic functions. The fitness value is calculated on a random sample of tasks that is generated in Sampling module. For each task, a random correct substitution for the pair (input set, output set) is selected, and the generated values for the estimated backdoor and output set variables are substituted into the original formula.

The behavior of this implementation is controlled by the following parameters:

• budget – An instance of the Budget module.

• measure – An instance of the Measure module.

• min_solved – The proportion of tasks in from the random sample for which a resolving substitution of values must be found. Default: 0.

Note

For this evaluation function to work correctly, you must specify the value of the budget parameter for the Measure module.

from function.impl import InverseBackdoorSets

function = InverseBackdoorSets(
    budget: Budget
    measure: Measure
    min_solved: float
)

Rho Function

Implementation of the fitness function from [AAAI2022]. This implementation is based on Guess-and-Determine function, however, the solver is used only in Unit Propagation mode.

The behavior of this implementation is controlled by the following parameters:

• measure – An instance of the Measure module.

• penalty_power – The multiplier for the penalty component in fitness value.

from function.impl import RhoFunction

function = RhoFunction(
    measure: Measure
    penalty_power: float
    only_propagate: bool
)

Inverse Polynomial Sets

This implementation is based on Inverse Backdoor Sets function, …. The behavior of this implementation is controlled by the following parameters:

• measure – An instance of the Measure module.

• min_solved – The proportion of tasks in from the random sample for which a resolving substitution of values must be found. Default: 0.

from function.impl import InversePolynomialSets

function = InversePolynomialSets(
    measure: Measure
    min_solved: float
    only_propagate: bool
)

Function modules

• Measure