1. 中国科学院长春光学精密机械与物理研究所应用光学国家重点实验室,吉林,长春,130033
2. 中国科学院长春光学精密机械与物理研究所,吉林,长春,130033
3. 中国科学院长春光学精密机械与物理研究所应用光学国家重点实验室,吉林,长春,130033
4. 中国科学院长春光学精密机械与物理研究所,吉林,长春,130033
纸质出版:2015
移动端阅览
李佩玥, 石俊霞, 郭嘉亮, 等. 一种混沌伪随机序列均匀化普适算法的改进[J]. 电子学报, 2015,43(4):753-759.
LI Pei-yue, SHI Jun-xia, GUO Jia-liang, et al. Improvement of a Universal Algorithm for Uniformization of Chaotic Pseudo-Random Sequences[J]. Acta Electronica Sinica, 2015, 43(4): 753-759.
李佩玥, 石俊霞, 郭嘉亮, 等. 一种混沌伪随机序列均匀化普适算法的改进[J]. 电子学报, 2015,43(4):753-759. DOI: 10.3969/j.issn.0372-2112.2015.04.018.
LI Pei-yue, SHI Jun-xia, GUO Jia-liang, et al. Improvement of a Universal Algorithm for Uniformization of Chaotic Pseudo-Random Sequences[J]. Acta Electronica Sinica, 2015, 43(4): 753-759. DOI: 10.3969/j.issn.0372-2112.2015.04.018.
为了分析盛利元等所述算法的安全性与普适性
从信息论的角度提出了单轮迭代信息损失量和动力学系统平均信息损失速度的概念
分析结果表明
第二类比特位变换的单轮迭代信息损失量为12比特
标准第二类比特位变换的单轮迭代信息损失量与指数
e
有关
存在信息损失量较小的可能性
将1023-
e
作为移位位数
使得标准第二类比特位变换无法遍历[-1
1
]
区间内的所有浮点数.进一步提出了暂态数据和第一类暂态变换的概念
并对文献[14
]
中所述算法进行了改进
改进后算法能够将任意混沌输出序列转换为至[0
1
]
区间内的浮点数
转换过程的信息损失量为
L
-1比特
接近有限计算精度为
L
时的最大信息损失速度
I
max
=
L
且通过
检验可证明转换后的混沌输出序列服从均匀分布.
In order to analyze the security and universality of the arithmetic proposed by Sheng et al
the concept of information loss in single iteration and average speed of information loss in dynamic system is proposed based on information theory.It is shown that the information loss of the 2nd bit-operation transformation is 12 bits
and which of the standard 2nd bit-operation transformation is related to the exponent
e
.It is possible that the information loss of the 2nd bit-operation transformation is so small.Not all of the float number in[-1
1
]
can be traversed by the standard 2nd bit-opera
tion transformation just because the 1023-
e
is used as the shift number.The concept of transient data and 1st transient transformation is proposed further
and the arithmetic proposed in literature[14
]
is improved as well.The output sequence of random digital chaotic system can be transformed as float number in[0
1
]
by the improved arithmetic.The information loss of this transformation is
L
-1 bits
which is approached to the maximum speed of information loss
I
max
=
L
under the computing precision
L
.The transformed sequence is uniform distributed which can be proved by
-verification.
0
浏览量
2
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621