基于高斯混合模型的物联网通信链路异常检测

Anomaly Detection of Internet of Things Communication Links Based on Gaussian Mixture Model

物联网涉及各种类型的设备、传感器和通信技术,链路异常的表现也各不相同,对于链路异常检测带来了挑战,为了精准地检测出移动物联网通信链路异常,提高通信质量,提出一种基于高斯混合模型的物联网通信链路异常检测方法。通过离散小波变换中的Mallat分解算法对移动物联网通信链路接收到的信号矢量展开小波分解,重构通信链路信号,并对重构后链路信号中的干扰部分展开白化处理,实现移动物联网通信链路干扰抑制。构建高斯混合模型,通过EM算法对高斯混合模型展开参数估计,利用完成参数估计后的模型展开移动物联网通信链路异常检测。实验结果表明,所提方法可以有效抑制通信过程中的干扰,同时还可以得到高准确率和高效率的通信链路异常检测结果。

The Internet of Things involves various types of devices,sensors,and communication technologies,and the manifestations of link anomalies are also different,which poses challenges for link anomaly detection.In order to accurately detect mobile IoT communication link anomalies and improve communication quality,a Gaussian mixture model-based IoT communication link anomaly detection method is proposed.Firstly,the Mallat decomposition algorithm in discrete wavelet transform was used to decompose the signal vector received by the communication link of mobile Internet of Things into wavelets,thus reconstructing the signal in communication link.Then,the interference part in the reconstructed signal was whitened to achieve the interference suppression in the communication link.Moreover,a Gaussian mixture model was constructed.Meanwhile,the parameters of the model were estimated by EM algorithm.Finally,the model after parameter estimation was used to detect the abnormalities in the communication link of mobile Internet of Things.The experimental results show that the proposed method can effectively suppress interference during communication and obtain accurate and efficient results of abnormality detection in the communication link of mobile Internet of Things.