A Harvard computer science professor has announced he has discovered a way to make codes unbreakable.

Michael Rabin, Watson professor of computer science, working with doctoral student Yan Zong Ding, said he has found a way to send and receive messages that cannot be decoded, a discovery that could have profound implications for cryptology.

"All the currently existing encryption schemes are based on improvable assumptions," Rabin said. "As a consequence, it is in principle possible that people using a secret algorithm and a sufficiently powerful computer will be able to break a code."

Rabin said people often eavesdrop on conversations and capture and store the coded messages. They then read the messages once they have perfected their code-breaking methods--a technique that Rabin's system would thwart.

"This new encryption preserves the secrecy of messages indefinitely so that even an adversary with unlimited computing power and who is infinitely smart in code breaking cannot ever decode the message," he said.

Rabin's encryption method involves using an infinite stream of randomly generated, unstored characters to transmit messages. The sender and receiver of a message agree when to start reading signals carried in the stream and the key to decode them, but the stream itself just disappears, leaving hackers no way to store and decipher the message.

Richard Lipton, a computer science professor at Princeton University, said he supports Rabin's claims.

"Rabin's method is new and potentially changes the crypto landscape," he said. "His work is very practical and is a major breakthrough."