基于动态模型的NTRU算法

Algorithm of dynamic patterns for NTRU

提出一种新的方法来改进NTRU算法执行速度。分析NTRU算法中多项式系数可能存在重复出现"11""101"等模型的分布特征,然后用贪心算法找出在多项式卷积计算时可以重复使用最多次数的模型,过滤多项式系数对模型的干扰,从而实现在多项式中发现模型数最大化。重复使用模型相应的卷积值,可以提高NTRU算法的密钥产生、加密和解密的速度。

This paper presented a new method to enhance the executive speed of NTRU algorithm. First, analyzed the polynomial coefficients to find out the distribution characteristics of patterns, such as ＂11＂＂ 101＂ and so on, which were possible to occur repetitively. Then, using greedy algorithm determined the patterns that could be used most frequently when calculating the convolution of polynomial, and eliminating the interferential effects that polynomial coefficients would have on patterns, thereby ultimately maximized the number of patterns found in the polynomial. By using the convolution of related patterns repetitively, speeds up the key generation, as well as the encryption and decryption operation of NTRU.