基于HNBJSON的物联网数据压缩方法  

Data compression about Internet of Things based on HNBJSON

作　　者：刘凯 钟永彦[1] 陈娟[1] 朱震 LIU Kai;ZHONG Yongyan;CHEN Juan;ZHU Zhen(School of Electrical Engineering,Nantong University,Nantong 226019,China)

机构地区：[1]南通大学电气工程学院,江苏南通226019

出　　处：《南京邮电大学学报（自然科学版）》2021年第6期29-34,共6页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家自然科学基金(62073154)资助项目。

摘　　要：针对物联网中因通信数据量大而导致的时延问题,提出一种基于HNBJSON的数据压缩方法。在云端使用朴素贝叶斯分类器对实时JSON数据进行分类,再按哈夫曼编码原理将数据生成压缩字典;基于压缩率更新压缩字典,并将压缩字典同步至边缘设备,用于数据压缩。将该方法应用于某高校3个不同教学场景的电力监测,实验结果表明,每条数据平均压缩时长约为0.98 ms,压缩率为76.9%左右,实现了边缘侧数据自适应无损压缩,具有较高的应用价值。Aiming at the problem of time delay caused by big data in the,Internet of Things(IoT),a data compression method based on Huffman⁃Naive Bayesian JSON(HNBJSON)is proposed.The real⁃time JSON data in the cloud with the Naive Bayesian principal are classified and the compressed dictionary is generated according to Huffman coding principle.To conduct data compression,the compression dictionary is updated based on the compression rate and synchronized to the edge device.In a college,the HNBJSON compression method is applied in the power monitoring in three scenarios.Experimental results show that each data transmission time can be reduced to 0.98ms with the method,and the compression rate can reach 76.9%.Obviously,the method can realized the adaptive lossless compression of communication data on the edge side.The method has a high application value.

关 键 词：时延 边云协同 字典压缩 朴素贝叶斯 哈夫曼编码 自适应 

分 类 号：TN919.3[电子电信—通信与信息系统]