About Random Number Production Algorithm Research

With computer network rapid development,the people more and more rely on in internet this open style platform carry on the information the exchange and the transmission.How both can use the interact quickly, effectively,and can guarantee transmits the data security,the secret, always are the academic circles important research directions.With unidirectional function and asymmetrical key theory appearance,in method structure safer encryption,but these methods validity is decided to a great extent by the random number quality height which uses.Random number the status which in the cryptology,the information study and so on in many application domains occupies more and more is also important.This article combine with some basic theory about random numbers, stochastic processes and random,research their principles of pseudo-random number algorithms and true random number generator, and analysis of the two types method of the advantages and disadvantages. For practical applications,this article proposes an algorithm which combination the advantages of pseudo-random number algorithms and true random number generator.This algorithm base on entropy of information theory.In the process of the algorithm implementation,by choosing a new generation of physical parameters to improve the entropy of system,this research uses a new dynamic data structure-- the Pool of Dynamic Buffed Data PDBD,through reading and writing Protection mechanisms to ensure data security.This article also improves the algorithm,and introduces data saved mechanism,buchang of chaotic,initial value systems and parallel algorithm.Through more the one unrelated systems running in parallel, the quality of generated random sequence has been greatly improved. Experimental tests show that this method can take into account the safety and speed,suitable for the actual application needs.