Font Size: a A A

Research And Design Of The Random Number Generator

Posted on:2013-01-24Degree:MasterType:Thesis
Country:ChinaCandidate:C L ShenFull Text:PDF
GTID:2218330371457616Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
The random number sequence has been widely used in scientific research, engineering and technology. With the more and more important role it plays, the design of the Random Number Generator becomes a research focus.After the introduction of the concept of randomness, the current research of the RNG is described. To be aimed at the problems of the True Random Number Generator and the Pseudo, a design of the Hybrid Random Number Generator is put forward which means that the seed random number is produced by TRNG and a large number of random number sequences which meet the actual needs are generated by the PRNG using these seed random numbers.The designed TRNG mainly consists of a metastable oscillating circuit and a management system with the single chip microcomputer as the core. In order to reduce the correlation of the data, two independent sampling circuits whose cycle values are prime numbers are designed. What's more, a new data compensation algorithm has been designed to improve the uniformity of the output. After the debugging, fifteen experiments are implemented and in each one 25 000 sets of data are analyzed. The results show that the uniformity and randomness of the output are good.The PRNG is realized on PC. For solving the low speed of the traditional generator, a new parallel combination RNG algorithm is designed. Compared with the traditional serial algorithm, the new one's speed is improved with an acceleration ratio of 3 in quad-core environment.In addition, a simple communication protocol is set up to ensure the communication between the TRNG and PRNG. After the completion of the entire system, the Diehard Tests is used to test the output for four times, every time with 20 million sets of data, all passed and mutually independent.The above work shows that the output of the HRNG has a certain practical value because of its real randomness, low cost and high speed.
Keywords/Search Tags:randomness, Hybrid Random Number Generator, oscillator sampling circuit, metastable phenomenon, parallel computing
PDF Full Text Request
Related items