Multi-layer cellular automata for generating normal random numbers

In this paper concept of multi-layer cellular automata and a novel neighborhood structure are introduced. According to these concepts, a novel approach for generating normal random numbers is proposed. First layer consists of binary cellular automata which are responsible for activating and inactivating cells in next layers. A cellular automaton with integer values is used for these layers. Interaction between layers of represented cellular automata leads to a dynamic and complex behavior of proposed model. Main idea of this model is based on central limit theorem to generate normal random numbers. To evaluate the quality of proposed model, several simulations are implemented. Results prove that multi-layer cellular automata generate better normal random numbers in comparison to MATLAB.