ID	原文	译文
5164	针对属于同一个家族的恶意软件的行为特征具有逻辑相似性这一特点，从行为检测的角度通过追踪 API函数调用的逻辑规则来提取恶意软件的特征，并利用静态分析与动态分析相结合的方法来分析恶意行为特征。	Aiming at the logical similarity of the behavioral characteristics of malware belonging to the same family, thecharacteristics of malware were extracted by tracking the logic rules of API function call from the perspective of behaviordetection, and the static analysis and dynamic analysis methods were combined to analyze malicious behavior characte-ristics.
5165	此外，依据恶意软件家族的目的性、继承性与多样性，构建了恶意软件家族的传递闭包关系，并改进了基于高斯混合模型的增量聚类方法来识别恶意软件家族。	In addition, according to the purpose, inheritance and diversity of the malware family, the transitive closure rela-tionship of the malware family was constructed, and then the incremental clustering method based on Gaussian mixturemodel was improved to identify the malware family.
5166	实验证明，所提方法不仅能节省恶意软件检测的存储空间，还能显著提高检测的准确率与识别率。	Experiments show that the proposed method can not only save thestorage space of malware detection, but also significantly improve the detection accuracy and recognition efficiency.
5167	现有的互联网面对激增的视频流量和工业机器应用，存在着大量的拥塞崩溃、数据分组时延等问题，	The current Internet faces a huge increase in video traffic and industrial machine applications, causing a lot ofcongestion and packet delays.
5168	而许多网络应用，例如工业互联网、远程医疗、无人驾驶、VR 游戏等，需要达到 1~10 ms 时延，微秒级抖动，但传统的网络只能将端到端的时延减少到几十毫秒。	Besides, many network applications, such as industrial Internet, telemedicine, driverless,VR games, etc, require a latency from 1 to 10 milliseconds and jitters within microsecond. But traditional network canonly reduce the end-to-end latency to tens of milliseconds.
5169	在这样的背景下，对网络端到端时延的控制如何从“尽力而为”到“准时、准确”，成为当前全球关注的热点领域。	Under this background, how to transit from “best effort” to“punctuality, accuracy”, controlling the end-to-end delay has become a hot field of global network research.
5170	通过介绍确定性网络的应用场景与需求，描述当前该领域的主要研究成果，总结分析了该领域的研究发展趋势和核心问题，期望对该领域的研究起到参考和帮助作用。	The applica-tion scenarios and requirements of the deterministic network were introduced, and the main research results of the currentparties were described as well as the trend of development and core issues were summarized. It hopes to provide a refer-ence and help for the research in this field.
5171	一个理想的代理重加密方案通常具有单向性、非交互性、可重复性、可控性和可验证性，然而目前的方案普遍只满足其中的 2 个或 3 个，在一定程度上降低了实用性。	A ideal proxy re-encryption scheme has five features, such as one-way encryption, non-interaction, repeatabil-ity, controllability and verifiability. The existing schemes, however, have only two or three of the five features, which re-duces the utility of them to some extent.
5172	为此，提出了一种支持 5 种特性的密文策略基于属性代理重加密（CP-ABPRE）方案。	For this, a new ciphertext-policy attribute-based proxy re-encryption(CP-ABPRE) scheme with the above five features was proposed.
5173	在所提方案中，云代理服务器只能利用重加密密钥重加密委托者指定的密文，抵御了满足重加密共享策略的用户与代理之间的共谋攻击；	In the proposed scheme, the cloud proxy server could only re-encrypt the ciphertext specified by the delegator by using the re-encryption key, and resist the collusion attack between the user and the proxy satisfying the re-encryption sharing policy.