Hardware Random number Generator for cryptography

One of the key requirement of many schemes is that of random numbers. Sequence of random numbers are used at several stages of a standard cryptographic protocol. A simple example is of a Vernam cipher, where a string of random numbers is added to massage string to generate the encrypted code. It is represented as $C=M \oplus K $ where $M$ is the message, $K$ is the key and $C$ is the ciphertext. It has been mathematically shown that this simple scheme is unbreakable is key K as long as M and is used only once. For a good cryptosystem, the security of the cryptosystem is not be based on keeping the algorithm secret but solely on keeping the key secret. The quality and unpredictability of secret data is critical to securing communication by modern cryptographic techniques. Generation of such data for cryptographic purposes typically requires an unpredictable physical source of random data. In this manuscript, we present studies of three different methods for producing random number. We have tested them by studying its frequency, correlation as well as using the test suit from NIST.