Stream Cipher Method Based on Neural Network
Haiyan Long · 2012
A new stream cipher algorithm based on one-way function of neural network is proposed.The nonlinear system comprised of such neural network is used to transform a group of LFSRs for key stream generation.As a result, the complexity of sequences is increased with period expanding.The produced cipher sequence has high randomness and has passed the standard tests of SP800-22.The experimental results show that the proposed encryption method is safe and has a higher performance of encryption and decryption speed by parallel computation structure.It can meet the need of stream cipher technology.The work has some values for deeper research on its theory and hardware application on secret communication.It is expected to attract more researchers in this field.