摘要
Random numbers play an increasingly important role in secure wire and wireless communication. Thus the design quality of random number generator(RNG) is significant in information security. A novel pseudo RNG is proposed for improving the security of network communication. The back propagation neural network(BPNN) is nonlinear, which can be used to improve the traditional RNG. The novel pseudo RNG is based on BPNN techniques. The result of test suites standardized by the U.S shows that the RNG can satisfy the security of communication.
Random numbers play an increasingly important role in secure wire and wireless communication. Thus the design quality of random number generator（RNG） is significant in information security. A novel pseudo RNG is proposed for improving the security of network communication. The back propagation neural network（BPNN） is nonlinear, which can be used to improve the traditional RNG. The novel pseudo RNG is based on BPNN techniques. The result of test suites standardized by the U.S shows that theRNGcan satisfy the security of communication.
基金
National Natural Science Foundation of China(60363087 ,90104005 and 60473023)
