摘要
对物联网多层设备信息通信数据进行分类识别,能够有效提高物联网设备的利用率,对设备信息通信数据进行识别,需要将通信信号特征从频域扩展至复平面,采用训练好的SVM对剩余通信数据进行分类识别。传统方法设定网络设备通信数据流的分形谱,推导出通信数据流的估计谱,但忽略了将通信信号特征扩展至复平面,导致识别误差大等问题。提出物联网多层设备信息通信数据分类识别方法,通过多层设备源信号的对角切片双谱提取通信信号特征,利用Chirp-Z变换将通信信号特征从频域扩展至复平面,选取出具有较强可分离度特征作为多层设备通信数据特征参数,并利用模糊C均值聚类算法进行初步聚类,选取聚类性能较优的通信数据作为支持向量机(SVM)的训练样本,再采用训练好的SVM对剩余通信数据进行分类识别。实验结果表明,所提方法有效提高了数据分类识别效率,且识别误差较小。
Traditional methods ignore the communication signal feature extended to the complex plane,which leads to large recognition error. Therefore,a method of classification and recognition for multi-layer equipment information communication data in the Internet of things was presented. At first,the multi-layer equipment source signal was used to extract the characteristic of communication signal from bispectral diagonal slice,and then Chirp-Z transform was used to extend communication signal feature from the frequency domain to the complex plane. Moreover,the feature with strong separable degree was selected as feature parameter of multi-layer equipment communication data.Meanwhile,the fuzzy C means clustering algorithm was applied to the preliminary clustering. Finally,communication data with good clustering performance was used as training sample of support vector machine( SVM),and then the trained SVM was used to classify and recognize the remaining communication data. Simulation results show that the proposed method effectively improves the efficiency of data classification and recognition. Meanwhile,the recognition error is small.
作者
魏葆春
甘发旺
WEI Bao-chun;GAN Fa-wang(Gansu Medical College,Pingliang Gansu 744000,China)
出处
《计算机仿真》
北大核心
2019年第1期425-428,436,共5页
Computer Simulation
关键词
物联网
多层设备
通信
分类识别
支持向量机
Internet of things
Multi-layer equipment
Communication
Classification and recognition
Support vector machine
