基于程序数据属性的联合软件特征技术研究

发布时间：2018-03-21 19:11

本文选题：软件盗版　切入点：软件特征　出处：《郑州大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着信息化程度的不断提高,信息技术的应用范围越来越广泛。人们在享受信息化便利的同时,也面临着信息技术所带来的安全问题。由于软件本身具有易复制、易篡改和易散布等特点,加之经济利益的驱动,软件盗版已经成为软件安全所面临的最严重的问题之一。软件特征技术是为了保护软件知识产权发展起来的新方向,已经成为软件保护的重要技术手段。软件特征是程序本身所固有的一个特性或多个特性的集合,软件特征技术通过提取、比较从不同软件中提取出来的固有特征,刻画其相似程度,以此辨别盗版。本文分析了软件特征技术的研究现状,阐述了软件特征的相关概念及分类,比较了不同软件特征技术各自的优缺点,并确定以动态和静态相结合的联合软件特征作为研究重点。程序的执行过程实质上是程序对数据进行处理的过程,在数据处理的过程中数据的调用序列以及值变化的序列和程序的特征是紧密相关的。本文提取程序的数据属性作为其特征可以很好地表征该程序独一无二的特性,使其区分于其他软件的能力得到提高。本文所做的主要工作有： 1、从静态和动态两个方面着手对程序的数据属性——常量和变量进行分析。对于常量,通过分析程序在特定输入下的动态执行过程,得到基于程序常量属性的子特征。对于变量,分别从静态和动态两个方面着手,分别得到基于程序变量属性的静态子特征和动态子特征。针对每一个子特征,提出合适的相似度计算方法。最后将三个子特征结合在一起,得到基于程序数据属性的联合软件特征。 3、通过实验评估基于程序数据属性的联合软件特征的性能,评估内容分为可信性评估和鲁棒性评估。其中,可信性评估从误判性、独立性和相似性三个方面来进行。本文还从宏观角度进行统计性比较试验,将本文提出的软件特征与现有的两个经典特征WPP和TaNaMM进行对比来验证本文特征的性能优缺点。理论和实验结果均表明：该特征具有较高的可信性和鲁棒性。
[Abstract]:With the continuous improvement of information technology, the application of information technology is becoming more and more extensive. While enjoying the convenience of information technology, people are also faced with the security problems brought by information technology. Because the software itself is easy to copy, Software piracy has become one of the most serious problems in software security. Software feature technology is a new direction in order to protect software intellectual property and has become an important technical means of software protection. Software feature is a collection of one or more characteristics inherent in the program itself. Software feature technology compares the inherent features extracted from different software and depicts the degree of similarity so as to identify piracy. In this paper, the research status of software feature technology is analyzed, the related concepts and classification of software feature are expounded, and the advantages and disadvantages of different software feature technologies are compared. It is determined that the joint software features, which combine dynamic and static state, are the key points of the research. The execution process of the program is essentially the process of processing the data by the program. In the process of data processing, the calling sequence of data and the sequence of value change are closely related to the characteristics of the program. The ability to distinguish it from other software has been improved. The main work done in this paper is:. 1. From the static and dynamic aspects, we analyze the data attribute of the program, constant and variable. For the constant, by analyzing the dynamic execution process of the program under the specific input, we get the subcharacteristics based on the program constant attribute. From the static and dynamic aspects, the static subfeatures and dynamic subfeatures based on the attributes of program variables are obtained respectively. For each subfeature, an appropriate similarity calculation method is proposed. Finally, the three sub-features are combined together. The joint software features based on program data attributes are obtained. 3. The performance of joint software features based on program data attributes is evaluated experimentally, and the evaluation content is divided into credibility evaluation and robustness evaluation. Independence and similarity. This paper also carries out statistical comparative experiments from a macro point of view. The software features proposed in this paper are compared with the existing two classical features, WPP and TaNaMM, to verify the performance advantages and disadvantages of this feature. The theoretical and experimental results show that the feature has high credibility and robustness.
【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.53

【参考文献】