基于时间自动机的物联网入侵检测方法  

Intrusion Detection Method of Internet of Things Based on Timed Automata

作　　者：钟小莉[1] 谢旻旻[1] 封希媛[1] ZHONG Xiao-li;XIE Min-min;FENG Xi-yuan(Computer School,Qinghai Nationalities University,Xining Qinghai 810007,China)

机构地区：[1]青海民族大学计算机学院,青海西宁810007

出　　处：《计算机仿真》2020年第10期278-281,356,共5页Computer Simulation

基　　金：基于立体图像智能分割技术应用的算法研究(2018XJY01)。

摘　　要：针对传统物联网入侵检测精度较低,且检测速度较慢的问题,提出基于时间自动机的物联网入侵检测方法。利用多个并发时间自动机构成一个时间自动机网络,采用共享变量法来完成同步传值通信,通过输出方给出一个特定共享变量赋值,输入方可以直接访问此共享变量,据此构建时间自动机网络模型。攻击者在入侵时,系统的响应时机以及攻击时机,可以反映出攻击者是否入侵成功,因此利用统计历史数据或者是实验的攻击计算行为评估攻击者权限,利用时间自动机叙述网络攻防对抗时的动态迁移特性,完成入侵检测。实验证明,所提方法检测精度较高,检测速度较快,具有一定的实用性。Due to low accuracy and slow detection speed of traditional IoT intrusion detection, a method to detect IoT intrusion based on time automaton was proposed. Many concurrent timed automata constituted a time automaton network. The shared variable method was used to complete the synchronous value-passing communication. The output side gave a specific shared variable assignment, and the input side directly accessed this shared variable. On this basis, the network model of timed automata was constructed. When an attacker made an incursion, the moment of response and attack could reflect whether the attacker had successfully invaded. Therefore, statistical historical data or simulation attack calculation behavior was used to evaluate the attacker’s authority. Finally, the time automaton was used to describe the dynamic migration feature during the network attack and defense. Thus, the intrusion detection was completed. Experiments show that the proposed method has higher detection accuracy and faster detection speed, so its practicality is high.

关 键 词：时间自动机 脆弱性 物联网 入侵检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]