Date Added: May 2012
The issue of random number generation is crucial for the implementation of cryptographic systems. Random numbers are often used in key generation processes, authentication protocols, zero-knowledge protocols, padding, in many digital signature and encryption schemes, and even in some side channel attack countermeasures. For these applications, security depends to a great extent on the quality of the source of randomness and on the way this source is exploited. The quality of the generated numbers is checked by statistical tests. In addition to the good statistical properties of the obtained numbers, the output of the generator used in cryptography must be unpredictable.