Tham khảo tài liệu 'hacker professional ebook part 180', công nghệ thông tin, kỹ thuật lập trình phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | From the Latin for one kind. Sometimes used to describe functions with a single argument such as the unary - the minus-sign as opposed to subtraction which presumably would be binary and that could get very confusing very fast. Thus monadic may be a better choice. Also see binary and dyadic. Unexpected Distance The values computed by a fast Walsh transform when calculating Boolean function nonlinearity as often used in S-box analysis. Given any two random Boolean sequences of the same length we expect to find about half of the bits the same and about half different. This means that the expected Hamming distance between two sequences is half their length. With respect to Boolean function nonlinearity the expected distance is not only what we expect it is also the best we can possibly do because each affine Boolean function comes in both complemented and uncomplemented versions. So if more than half the bits differ between a random function and one version then less than half must differ to the other version. This makes the expected distance the ideal reference point for nonlinearity. Since the FWT automatically produces the difference between the expected distance and the distance to each possible affine Boolean function of the given length I call this the unexpected distance. Each term is positive or negative depending on which version is more correlated to the given sequence and the absolute value of this is a measure of linearity. But since we generally want nonlinearity we typically subtract the unexpected value from half the length of the sequence. Unicity Distance The amount of ciphertext needed to uniquely identify the correct key and its associated plaintext assuming a ciphertext-only attack and natural language plaintext . With less ciphertext than the unicity distance multiple keys may produce decipherings which are each plausible messages although only one of these would be the correct solution. As we increase the amount of ciphertext many .
