05.12.2012 Views

NASA Scientific and Technical Aerospace Reports

NASA Scientific and Technical Aerospace Reports

NASA Scientific and Technical Aerospace Reports

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

using a value of higher order differential which holds with probability. The success of attack depends on the probability,<br />

however, the necessary number of chosen paintext <strong>and</strong> computational cost become very small. ICEBERG is a block cipher<br />

with sixteen round SPN structure. In this paper, we analyze its higher order differential property, <strong>and</strong> estimate its strength<br />

against higher order differential attacks. As a result, we found that five round ICEBERG is attackable using eighth order<br />

differential with 2,304 chosen cipher texts <strong>and</strong> 2(sup 85) times round function calculations. And in the case using probabilistic<br />

seventh order differential, it is attackable with 1,152 chosen cipher texts <strong>and</strong> 283 times round function calculations <strong>and</strong><br />

probability about 0.7.<br />

Author<br />

Cryptography; Differential Equations; Algorithms; Reconfigurable Hardware<br />

20060002083 National Inst. of Information <strong>and</strong> Communications Technology, Tokyo, Japan<br />

Error Detection <strong>and</strong> Authentication in Quantum Key Distribution<br />

Yamamura, Akihiro; Ishizuka, Hirokazu; Journal of the National Institute of Information <strong>and</strong> Communications Technology.<br />

Special Issue on Information Security, Volume 52, Nos. 1/2; March/June 2005, pp. 141-151; In English; See also<br />

20060002073; Copyright; Avail.: Other Sources<br />

Detecting errors in a raw key <strong>and</strong> authenticating a private key are crucial for quantum key distribution schemes. Our aim<br />

is to propose practical methods for error detection <strong>and</strong> authentication in quantum key distribution schemes. We introduce<br />

several concepts about neighborhood collision free properties of Boolean functions, which are closely related to hash<br />

functions, <strong>and</strong> propose methods based on neighborhood collision free functions <strong>and</strong> error correcting codes such as<br />

Reed-Solomon code. We also examine whether or not widely used cryptographic hash functions SHA-1 <strong>and</strong> MD5 satisfy the<br />

neighborhood collision free property by computation experiments.<br />

Author<br />

Computer Information Security; Error Correcting Codes; Quantum Cryptography<br />

20060002084 National Inst. of Information <strong>and</strong> Communications Technology, Tokyo, Japan<br />

Secret Sharing Scheme Using Natural Language Text<br />

Takizawa, Osamu; Yamamura, Akihiro; Makino, Kyoko; Journal of the National Institute of Information <strong>and</strong> Communications<br />

Technology. Special Issue on Information Security, Volume 52, Nos. 1/2; March/June 2005, pp. 173-183; In English; See also<br />

20060002073; Copyright; Avail.: Other Sources<br />

Modifying the idea of the visual cryptography, we propose a method of sharing a secret key using natural language tests.<br />

Our target here is restricted to Japanese tests. Each participant obtains a share, which is a Japanese test in our scheme. When<br />

a certain number of participants retrieve the secret key, they supply their shares <strong>and</strong> pile up these natural language tests. The<br />

sequence of the first, second (<strong>and</strong> so on) letters occurred in the pile shows the secret text. The order of the pile is significant,<br />

<strong>and</strong> changing the order may yield the distinct secret text. It is easy to pile the shared natural language texts by computer<br />

operation. Human eyes can recognize the secret text from the piled texts. However, we aim to construct a natural language<br />

text secret sharing scheme employing a morphological analyzer because a meaningless phrase is a chain of morphemes<br />

consisting of one word with a high probability. We can make a shared natural text look like a natural text without any secret<br />

meaning by synthesizing using a text database.<br />

Author<br />

Cryptography; Natural Language (Computers); Algorithms; Texts<br />

20060002089 National Inst. of Information <strong>and</strong> Communications Technology, Tokyo, Japan<br />

A Generalized Harley Algorithm for Genus Two Hyperelliptic Curves<br />

Sugizaki, Hiroki; Kazuto, Matsuo; Chao, Jinhun; Tsujii, Shigeo; Journal of the National Institute of Information <strong>and</strong><br />

Communications Technology. Special Issue on Information Security, Volume 52, Nos. 1/2; March/June 2005, pp. 111-117; In<br />

English; See also 20060002073; Copyright; Avail.: Other Sources<br />

A fast addition algorithm for divisor classes of genus two hyperelliptic curves over finite fields of odd characteristics was<br />

proposed by Harley in 2000 <strong>and</strong> a lot of improvements of the algorithm has been proposed, besides extensions of the algorithm<br />

for the curves over finite fields of characteristic two have been proposed by the authors <strong>and</strong> Lange independently. However,<br />

any Harley algorithm over arbitrary characteristic firlds have not been known until now. This paper shows a generalization<br />

of the Harley algorithm to genus two hyper elliptic curves over finite fields of arbitrary characteristics. The proposed algorithm<br />

169

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!