摘要
针对以前的保持前缀地址随机化算法或是性能不高,或是不能进行一致的随机化的问题,该文提出了一种基于随机字符串的保持前缀地址随机化算法。该算法使用预先得到的随机字符串,避免了计算的开销,加快了地址随机化的速度。使用基于随机字符串算法和C rypto-pan的混合算法,可以进一步避免大的内存开销。实验结果表明,对相同的需要随机化的地址数,基于随机字符串的算法使用的时间只相当于C rypto-pan的1/10。采用二者的混合算法,一般使用的时间相当于C rypto-pan的1/4,而所使用的内存要比基于随机字符串所使用内存少得多,因此能够实现更好的时间、空间折衷。
Prefix preserving anonymization schemes are either slow or inconsistent. This paper describes a random bit string-based consistent prefix-preserving IP address anonymization scheme, which uses a pre-computed random bit string to reduce the computational overhead and accelerate the process. The hybrid algorithm with crypto-pan further reduces the large memory requirement. Experiments indicate that the time required for the bit string-based algorithm to process a set of addresses is only 1/10 that of the Crypto-pan algorithm. The time required for the hybrid algorithm is only 1/4 that of the Crypto-pan algorithm with far less memory than the bit string based approach. Therefore, the hybrid algorithm provides a better memory time compromise.
出处
《清华大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2006年第10期1723-1726,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家"九七三"高技术项目(2003CB314807)
关键词
IP地址随机化
流量采集
流量隐私去除
流量工程
IP address anonymization
traffic caplure
traffic anonymization
traffic engineering