Category: Science & Technology Posted: November 4, 2014 11:04AM Author: Guest_Jim_* <\/p>\n
Data security is of great importance in the modern world, with so much private information being transmitted every day. Many modern cryptographic schemes rely on such complex operations that massive computational resources would be needed to crack them. Researchers at MIT and Maynooth University have recently found that some of these schemes also allow for another kind of security, which could be used to better protect the data. <\/p>\n
While computational complexity protects data by requiring prodigious amounts of time and resources to crack, information-theoretic security protects against extracting any useful information, even with unlimited computational power. Information-theoretic schemes are some complicated though, that they are not practical to use. However, the researchers have found that existing cryptographic schemes do have some information-theoretic guarantees. The researchers found this by examining the probability spaces that pieces of plaintext would become certain ciphertexts. By keeping the probability of the different ciphertexts near each other, it becomes much harder to infer what the original plaintext may be. <\/p>\n
What this translates to is that while an entire message may not be information-theoretic secured, portions of it could be. Potentially one could design the scheme to make sure select portions are, such as social security numbers, ages, or other particularly important pieces of information. <\/p>\n
Source: MIT <\/p>\n
<\/p>\n