恶意软件通过搜索引擎扩散行为仿真与控制研究

发布时间：2018-06-14 15:13

本文选题：无标度网络 + 恶意软件扩散　；参考：《华中科技大学》2012年硕士论文

【摘要】：互联网信息数量的急速增长，使得搜索引擎的作用越来越重要，进而使得越来越多的恶意软件选择通过搜索引擎进行传播。本文通过研究恶意软件经由搜索引擎在网络中进行传播的行为过程，建立了恶意软件扩散的多智能体仿真模型，并针对仿真模型得出的结果，提出了几条具有参考性的控制策略。首先，，本文针对用户通过搜索引擎点击网站后，主机在恶意软件扩散过程中所处的不同状态，将复杂网络中的主机分为五种：永久免疫主机、永久感染主机、易感染主机、暂时感染主机、暂时免疫主机。根据主机状态变化过程和状态变化影响因素，列出主机不同状态变化的数学公式。然后，本文将复杂网络技术和多智能体技术相结合，根据文章中的数学公式，设计了恶意软件在复杂网络中传播的多智能体仿真模型。该模型以无标度网络为研究环境，通过探讨永久免疫主机的数量、永久恶意主机的数量、恶意软件感染率、主机的学习能力和抵抗力、恶意软件的进化能力、恶意软件的传播方式、恶意软件的传播能力、用户的“口碑”效应等，来分析影响恶意软件扩散的相关因素。随后，根据恶意软件传播的仿真模拟结果和敏感性分析发现，永久免疫主机的数量、主机的学习能力对控制恶意软件的传播具有正相关性。永久恶意主机的数量、恶意软件的进化能力、恶意软件的传播率、用户的“口碑”效应对控制恶意软件的传播具有正相关性。恶意软件的传播方式对恶意软件传播也具有一定的影响。最后，针对恶意软件的利益链条、实验仿真结果和敏感性分析，提出几条遏制恶意软件传播行之有效的方法。主要从政府、网站、法律、用户等方面综合考虑：政府方面要提高其监管作用；网站方面搜索引擎要当好网络把关人，增加大型网站的数量，增加行业自律性；法律方面要制定完善的法律法规；用户方面要养成良好上网习惯，提高主机免疫力。
[Abstract]:With the rapid growth of Internet information, the role of search engines becomes more and more important, and more malware are selected to spread through search engines. In this paper, a multi-agent simulation model of malware diffusion is established by studying the behavior of malicious software propagating through the network through search engine, and the results obtained by the simulation model are discussed. Several control strategies with reference are proposed. First of all, aiming at the different status of the host in the process of malware diffusion after the user clicks on the website through the search engine, this paper divides the host in the complex network into five kinds: permanent immune host, permanent infection host, vulnerable host. Temporary infection host, temporary immune host. According to the process of host state change and the influencing factors of state change, the mathematical formulas of host state change are listed. Then, this paper combines the complex network technology with the multi-agent technology, and designs a multi-agent simulation model of malware spread in the complex network according to the mathematical formula in this paper. The model takes scale-free network as the research environment, and discusses the number of permanent immune hosts, the number of permanent malicious hosts, the infection rate of malware, the learning ability and resistance of hosts, and the evolutionary ability of malicious software. The spread mode of malware, the spreading ability of malware, the word-of-mouth effect of users and so on, are analyzed to analyze the related factors that affect the spread of malware. Then, according to the simulation results and sensitivity analysis of malware propagation, it is found that the number of permanent immune hosts and the learning ability of hosts have a positive correlation to control the spread of malware. The number of permanent malware hosts, the evolutionary ability of malware, the spread rate of malware and the "word-of-mouth" effect of users have positive correlation to control the spread of malware. Malware communication mode also has certain influence on malware communication. Finally, according to the benefit chain of malware, the simulation results and sensitivity analysis, several effective methods to curb the spread of malware are put forward. Mainly from the government, website, law, users and other aspects of comprehensive consideration: the government should improve its regulatory role, website search engine should be a good network gatekeeper, increase the number of large websites, increase the self-discipline of the industry; The law should make perfect laws and regulations, and users should develop good internet habits and improve host immunity.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP393.08;TP391.3

【参考文献】