摘要
针对内存-时间权衡法在哈希密码分析中存在的哈希链碰撞缺陷,给出了彩虹表和MD5算法相应的截短函数处理流程,并利用CUDA架构的GPU(图形处理器)单指令多线程特点在并行迭代运算和遍历哈希链上的优势,提出了一种基于GPU加速的彩虹表分析MD5哈希密码的方法。
Hellman' s time-memory trade-off method uses the idea of Brute Force and dictionary to get the time and space compromise. For reducing the collisions in the preprocess of hash chains, Oechslin designed the Rainbow Tables which use different reduction functions to solve the problem. With the advantage of GPU single instruction multiple threads in the iterative computation and parallel traversal hash chains, we proposed a new implementation of GPU in CUDA architecture accelerated Rainbow Tables analysis of MD5 hash password.
出处
《重庆理工大学学报(自然科学)》
CAS
2013年第7期61-66,共6页
Journal of Chongqing University of Technology:Natural Science
基金
教育部人文社会科学研究青年基金项目(10YJC870037)