物联网中海量数据管理技术研究

发布时间：2018-03-24 19:45

本文选题：物联网　切入点：海量数据　出处：《广东工业大学》2015年硕士论文

【摘要】：随着信息技术的不断发展,使得数据成为整个系统中非常关键的元素,目前物联网设备从周围物质世界获取的数据具有多源异构性、海量性、时空相关性和实时性等特点。针对物联网数据处理难点,目前的研究还存在着诸多局限,如何提供一个通用的数据管理平台、对数据进行有效管理实现快速准确地查询以及对内部数据访问安全性的管理等都是物联网数据管理中比较重要的问题。首先,本文分析了物联网海量数据管理的研究背景、意义和国内外当前发展现状。通过对物联网数据生命周期的分析,总结出其整个的生命周期,提出了一个数据管理框架来适用不同层次的物联网应用,并详细说明了该框架的分层架构,针对各层模块的功能进行了阐述。此外,考虑到物联网数据资源除了时空相关性之外还具有语义相关性,通过介绍了目前的时空索引和倒排索引的构成和基本要素,提出了一种语义IR树索引结构。该索引结构在IR索引树的基础上进行改进,在索引的构建过程中加入了数据的语义特性,以应对相关查询。文中描述了该索引的建立过程及对查询处理的优化,最后通过实验来与IR树和R树进行比较,验证了它的有效性,可一定程度上提高资源查找的准确率。最后,针对物联网局域子网内部数据访问的应用场景,本文提出了一种双要素认证会话机制,增加了会话ID参数来参与认证过程。在用户成功鉴权且在会话有效期内的情况下,用户可以不通过鉴权而直接与内传感器节点通信；在有效期或冷冻期的情况下,用户只需进行简单的验证则可以延长会话有效期。通过实现一次鉴权多次通信,从而减轻了每次重复鉴权的通信压力,在增加微小计算代价的基础上提高了数据访问的效率。本文主要研究物联网的数据管理,通过分析总结物联网数据的生命周期,提出了一个数据管理框架来支持不同层次的物联网应用；提出了一种语义IR树索引结构,可应用于同时具有时空相关性和语义相关性的资源,提高查找的准确率；提出了一种双要素认证会话机制在增加微小计算代价的基础上实现对物联网局域子网内部数据的安全高效访问。
[Abstract]:With the continuous development of information technology, data becomes a very key element in the whole system. At present, the data obtained from the material world around the Internet of things has multi-source heterogeneity and magnanimity. In view of the difficulties of data processing in the Internet of things, the current research still has many limitations, how to provide a general data management platform, Effective data management, fast and accurate query and internal data access security management are important issues in Internet of things data management. Firstly, this paper analyzes the research background of Internet of things mass data management. By analyzing the life cycle of the Internet of things data, the whole life cycle of the Internet of things is summarized, and a data management framework is proposed to apply to different levels of Internet of things applications. The hierarchical architecture of the framework is explained in detail, and the functions of each layer module are expounded. In addition, considering that the data resources of the Internet of things have semantic correlation in addition to the temporal and spatial correlation, By introducing the structure and basic elements of the present spatio-temporal index and inverted index, a semantic IR tree index structure is proposed, which is improved on the basis of IR index tree. In order to deal with the related queries, this paper describes the process of building the index and the optimization of query processing, and compares it with IR tree and R-tree through experiments. The validity of the method is verified, and the accuracy of resource search is improved to some extent. Finally, a two-element authentication session mechanism is proposed for the application of data access in the Internet of things local area network. A session ID parameter is added to participate in the authentication process. In the case of successful authentication of the user and within the validity period of the session, the user can communicate directly with the internal sensor node without authentication; in the case of validity or freezing period, A user can extend the validity of a session by simply verifying it. By implementing a single authenticated multiple communication, the communication pressure of each duplicate authentication is reduced. On the basis of increasing the cost of micro-computing, the efficiency of data access is improved. This paper mainly studies the data management of the Internet of things, and summarizes the life cycle of the Internet of things data by analyzing and summarizing the life cycle of the Internet of things data. A data management framework is proposed to support different levels of Internet of things applications, and a semantic IR tree index structure is proposed, which can be applied to resources with both temporal and spatial correlation and semantic correlation to improve the accuracy of searching. A two-element authentication session mechanism is proposed to secure and efficiently access the internal data of the Internet of things (IoT) local area network on the basis of increasing the cost of micro-computing.
【学位授予单位】：广东工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.44;TN929.5

【相似文献】