基于主动信任的边缘网络质量增强数据获取技术

海量的感知设备正以日益增长的速度接入边缘网络，构成了当前网络最主要的数据获取平台。数据的可信与高质量获取是信息技术领域的基石以及关键的科学问题之一。针对以往数据获取过程易受到各种攻击与非可信操作，导致获取的数据中存在不可信数据、数据质量低等问题，本课题研究基于主动信任获取的高质量数据获取机制与方法。首先建立数据可信与质量判别的度量机制与标准，给出基于可信基础上数据质量判别准则，用以量化选取的数据对满足应用需求的程度。其次构建一种新型的主动信任获取机制与方法，从而能够在复杂的边缘网络中快速的、主动的、可靠性高的获得评测对象的信任度。从数据获取的源头鉴别恶意与可信数据，从源头保证数据获取结果的安全可信。最后建立包括终端感知数据的质量评估，到边端协同的数据优选，边云协同数据融合的“端－边－云” 多层、多级、分布式高质量数据获取体系结构，实现基于主动信任的边缘网络数据获取框架。

Massive sensing devices are connecting to the edge network with acceleration, and constitutes the most important data acquisition platform of the current network. Credibility and high-quality access to data is the cornerstone and one of key scientific issues of information technology. In the past, the data acquisition process was susceptible to various attacks and untrusted operations, which led to the existence of untrusted data and low data quality. In response to these problems, this project studies high-quality data acquisition mechanisms and methods based on active trust acquisition. First, a measurement mechanism and standard for data credibility and quality discrimination are established, and data quality discrimination criteria based on credibility are given to quantify the degree to which the selected data meets application requirements. Then, a new active trust acquisition mechanism and method is constructed, which can quickly, actively and reliably obtain the trust degree of the evaluated object in the complex edge network, and identify malicious and trusted data from the source of the data acquisition, and ensure the data security and credibility. Finally, a “terminal-edge-cloud” multi-layer, distributed, and high-quality data acquisition architecture is established, including the quality evaluation of terminal-aware data, the data optimization selection based on edge-terminal collaboration, and the data fusion based on edge-cloud collaboration, so as to implement an active trust-based data acquisition framework for edge networks.