返回
Evolutionary hash functions for specific domains
DOI:10.1016/j.asoc.2019.02.014.png)
摘要
En 中文
Hash functions are a key component of many essential applications, ranging from compilers, databases or internet browsers to videogames or network devices. The same reduced set of functions are extensively used and have become standard de facto since they provide very efficient results in searches over unsorted sets. However, depending on the characteristics of the data being hashed, the overall performance of these non-cryptographic hash functions can vary dramatically, becoming a very common source of performance loss. Hash functions are difficult to design, they are extremely non-linear and counter-intuitive, and relationships among variables are often intricate and obscure. Surprisingly, very little scientific research is devoted to the design and experimental assessment of these widely used functions. In this work, in addition to performing an up-to-date comparison of state-of-the-art hash functions, we propose the use of evolutionary techniques for designing ad hoc non-cryptographic hash functions. Thus, genetic programming will be used to automatically design a tailor-made hash function that can be continuously evolved if needed, so that it is always adapted to real-world dynamic environments. To validate the proposed approach, we have compared several quality metrics for the generated functions and the most widely used non-cryptographic hash functions across eight different scenarios. The results of the evolved hash functions outperformed those of the non-cryptographic hash functions in most of the cases tested. (C) 2019 Elsevier B.V. All rights reserved.
Keyword:
Genetic programming
Hash functions
Evolutionary algorithm
Automated design
AI总结
对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
6.6
论文数:
1.4W
被引数:
4.8W
机构
引用论文
Open-Source Software Radio Performance for Cellular Communications Research with UAV Users面向无人机用户蜂窝通信研究的开源软件无线电性能

