当前位置: X-MOL 学术Arch. Computat. Methods Eng. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A Technical Survey on Intelligent Optimization Grouping Algorithms for Finite State Automata in Deep Packet Inspection
Archives of Computational Methods in Engineering ( IF 9.7 ) Pub Date : 2020-04-15 , DOI: 10.1007/s11831-020-09419-z
Prithi Samuel , Sumathi Subbaiyan , Balamurugan Balusamy , Sumathi Doraikannan , Amir H. Gandomi

Construction and deployment of finite state automata from the regular expressions might results in huge overhead and results in the state explosion problem which is in need of large memory space, high bandwidth and additional computational time. To overcome this problem, a new framework is proposed, and several intelligent optimization algorithms are reviewed and compared in this framework. The proposed approach is called intelligent optimization grouping algorithms (IOGA), which intends to group regular expression intelligently. IOGAs are used to allocate the regular expression sets into various groups and to build independent deterministic finite automata (DFA) for each group. Grouping the regular expression efficiently solves the state explosion problem by achieving large-scale best tradeoff among memory utilization and computational time. This study reviews and compares the various alternatives of IOGA including genetic algorithm, ant colony optimization, particle swarm optimization, bacterial foraging optimization, artificial bee colony algorithm, biogeography based optimization, cuckoo search, firefly algorithm, bat algorithm and flower pollination algorithm for solving the problem of DFA state explosion and also for improving the overall efficiency of deep packet inspection (DPI). The discussions state that by effectively using these grouping algorithms along with DFA based DPI, the number of states can be reduced, providing a balance between the memory consumption, time complexity, throughput, inspection speed, convergence speed and grouping time.



中文翻译:

深度数据包检测中有限状态自动机智能优化分组算法的技术综述

从正则表达式构造和部署有限状态自动机可能会导致巨大的开销,并导致状态爆炸问题,这需要较大的存储空间,高带宽和额外的计算时间。为了克服这个问题,提出了一个新的框架,并在该框架中对几种智能优化算法进行了回顾和比较。所提出的方法称为智能优化分组算法(IOGA),旨在对正则表达式进行智能分组。IOGA用于将正则表达式集分配到各个组中,并为每个组构建独立的确定性有限自动机(DFA)。通过对正则表达式进行分组,可以在内存利用率和计算时间之间实现大规模的最佳权衡,从而有效地解决了状态爆炸问题。本研究回顾并比较了IOGA的各种替代方案,包括遗传算法,蚁群优化,粒子群优化,细菌觅食优化,人工蜂群算法,基于生物地理的优化,布谷鸟搜索,萤火虫算法,蝙蝠算法和花朵授粉算法,以解决这些问题。 DFA状态爆炸的问题,还用于提高深度数据包检查(DPI)的整体效率。讨论指出,通过有效地使用这些分组算法​​以及基于DFA的DPI,可以减少状态数,从而在内存消耗,时间复杂度,吞吐量,检查速度,收敛速度和分组时间之间取得平衡。粒子群优化,细菌觅食优化,人工蜂群算法,基于生物地理学的优化,布谷鸟搜索,萤火虫算法,蝙蝠算法和花朵授粉算法,用于解决DFA状态爆炸问题,并提高深包检查(DPI)的整体效率)。讨论指出,通过有效地使用这些分组算法​​以及基于DFA的DPI,可以减少状态数,从而在内存消耗,时间复杂度,吞吐量,检查速度,收敛速度和分组时间之间取得平衡。粒子群优化,细菌觅食优化,人工蜂群算法,基于生物地理学的优化,布谷鸟搜索,萤火虫算法,蝙蝠算法和花授粉算法,用于解决DFA状态爆炸问题并提高深层数据包检查(DPI)的整体效率)。讨论指出,通过有效地使用这些分组算法​​以及基于DFA的DPI,可以减少状态数,从而在内存消耗,时间复杂度,吞吐量,检查速度,收敛速度和分组时间之间取得平衡。蝙蝠算法和花授粉算法解决了DFA状态爆炸问题,也提高了深度数据包检查(DPI)的整体效率。讨论指出,通过有效地使用这些分组算法​​以及基于DFA的DPI,可以减少状态数,从而在内存消耗,时间复杂度,吞吐量,检查速度,收敛速度和分组时间之间取得平衡。蝙蝠算法和花授粉算法解决了DFA状态爆炸问题,也提高了深度数据包检查(DPI)的整体效率。讨论指出,通过有效地使用这些分组算法​​以及基于DFA的DPI,可以减少状态数,从而在内存消耗,时间复杂度,吞吐量,检查速度,收敛速度和分组时间之间取得平衡。

更新日期:2020-04-20
down
wechat
bug