The block cipher for construction of S-boxes based on particle swarm optimization

Abstract

S-boxes block cipher is the only non-linear components, and its password strength decides the security of the entire block cipher strength. It is very difficult and complex to use the traditional methods of mathematical to construct the S-boxes with good performance. This thesis will put on a new construction of the s-boxes method based on the particle swarm optimization (PSO). This algorithm defines the location of the two particles “addition” operator, replacement of the original speed equation in order to achieve the exchange of information between particles, and the addition of a mutation operator to maintain the population diversity to prevent premature convergence. The experimental results show that it is plausible to use POS to get a group of S-boxes with high non-linearity and low evenness in difference. This structure can effectively reduce the redundant computation to accelerate the convergence speed.