基于物联网数据处理的关键技术研究

发布时间：2019-06-24 17:59

【摘要】：随着物联网的快速发展,加之互联网的普及,使各种物品接入互联网成为了可能,这给人们的日常生活带来了极大的便利,同时又极大地促进了社会经济的发展,但与此同时,物联网也给人们带来了许多难题。首先,由于物联网中将包含更多的设备和物品,它们将会产生更多的数据,所以现有的计算资源将受到挑战,如何部署计算资源来处理海量的数据是一大难题。而解决该问题的一个有效的方法就是利用分布式技术,可以预见,随着数据的增长,分布式技术将成为物联网时代的一项基本计算技术。其次,由于互联网的开放性,作为在其终端进行了扩展的物联网,势必也有一定的开放性,加之近年来网络诈骗频发,这不得不让人们对物联网时代的个人隐私问题感到担忧,所以对物联网安全技术进行相关的研究,可以确保物联网健康快速的发展,真正促进社会的发展,提高人们的生活水平。最后,在物联网时代,数据将呈爆炸性增长,将会产生海量的数据,而如何处理这些数据,从海量的数据中快速地查找到人们需要的信息,也将成为物联网时代需要解决的一个难题。所以本文对基于物联网有关数据处理问题进行重点研究。综上所述,本文主要研究了分布式技术,物联网信息安全技术和文本处理技术,具体包括分布式系统原理,彩虹表的生成技术,彩虹表的密码破解的加速技术,文本处理技术等。首先,本文对分布式系统的架构和原理进行了仔细的研究,并自行搭建了一个集群,以供后续相关实验使用；其次,本文认真学习并深入研究了彩虹表生成的原理,以及彩虹表的生成加速技术,并进行了相关彩虹表的生成实验,以及利用分布式集群进行了加速彩虹表的生成实验,并对基于彩虹表的密码破解技术进行了深入的研究,利用生成的彩虹表,进行了相关破解实验,并利用相关技术,设计并实现了一个基于彩虹表的word2003文档密码破解系统,并对word2003加密文档进行了相关破解实验,以及利用分布式来加速破解的实验,其中破解成功率高达99%；最后,本文对基于物联网数据处理的文本处理技术进行了深入的学习和研究,并在此基础上,设计并实现了一个基于LDA主题模型的文本分类系统,并利用相关训练数据,对其进行训练,以及进行了相关文本分类测试实验,分类结果正确率高达91%。
[Abstract]:With the rapid development of the Internet of Things and the popularization of the Internet, the access of various items into the Internet has become possible, which brings great convenience for people's daily life, and also greatly promotes the development of the social economy, but at the same time, The Internet of Things has also brought many challenges to people. First, since more devices and items will be included in the Internet of Things, more data will be generated, so existing computing resources will be challenged, and how to deploy computing resources to handle massive data is a great challenge. One effective way to solve this problem is to use the distributed technology, and it is expected that with the increase of the data, the distributed technology will become a basic computing technology of the Internet of Things. Second, because of the openness of the Internet, as an Internet of things to be expanded at its terminal, it is bound to have a certain openness, and in recent years, network fraud is frequent, which has to let people worry about the privacy of the Internet of things. Therefore, the research on the safety technology of the Internet of Things can ensure the rapid development of the Internet of Things, and truly promote the development of the society and improve the living standard of people. In the end, in the age of Internet of things, the data will be explosive, and the data will be generated, and how to handle the data can quickly find the information needed by people from the massive data, and it will become a problem that needs to be solved in the age of Internet of things. Therefore, this paper focuses on the data processing problem based on the Internet of things. In the light of the above, this paper mainly studies the technology of distributed technology, Internet of things information security and text processing technology, including the principle of distributed system, the generation technology of the rainbow table, the accelerated technology of the password cracking of the rainbow table, and the text processing technology. First of all, this paper makes a careful study of the architecture and principle of the distributed system, and set up a cluster to be used for subsequent relevant experiments. Secondly, this paper studies and deeply studies the principle of the rainbow table generation and the generation and acceleration technology of the rainbow table. In this paper, the generation experiment of the related rainbow table is carried out, and the generation experiment of the accelerated rainbow table is carried out by using the distributed cluster, and the password cracking technology based on the rainbow table is deeply researched, the related cracking experiment is carried out by using the generated rainbow table, and the related technology is utilized. In this paper, a word2003 document password cracking system based on the rainbow table is designed and implemented, and the related cracking experiment of the world d2003 encrypted document is carried out, and the experiment is accelerated by using the distributed algorithm, in which the success rate of the cracking is up to 99%, and the most In this paper, the text processing technology based on the Internet of Things data processing is studied and studied deeply, and on this basis, a text classification system based on the LDA subject model is designed and implemented, and the relevant training data is used to train the text classification system. practice, and carry out the relevant text classification test experiment, the correct rate of the classification result is as high as 91
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.44;TN929.5;TP391.1

【相似文献】