摘要
针对传统物联网深度包流量检测效率过低问题,提出一种基于物联网流量指纹的安全威胁轻量级检测方法。首先采用数据重构的方法获取流量时空数据,然后采用深度学习的方法提取流量数据时空特征(即流量数据指纹),最后采用基于蚁群算法优化的BP神经网络进行流量异常检测和识别。实验证明,使用该算法进行流量异常检测能够避免检测模型陷入局部最优,能够显著提高物联网威胁检测精度。
Aiming at the low efficiency of traditional deep packet traffic detection in Internet of Things(IoT),this paper proposes a lightweight detection method for security threat based on Internet of things traffic fingerprints.Firstly,the spatiotemporal data of traffic are obtained by data reconstruction method.Then,a deep learning method is used to extract the spatiotemporal characteristics of traffic data(i.e.,traffic data fingerprint).Finally,a BP neural network optimized by ant colony algorithm is used to detect and identify traffic anomaly.Experiments show that the proposed algorithm for traffic anomaly detection prevents the detection model from falling into a local optimum,and significantly improves the accuracy of IoT threat detection.
作者
赵研
ZHAO Yan(Digital Guangdong Network Construction Co.,Ltd.,Guangzhou 510030,China)
出处
《移动通信》
2021年第3期62-66,共5页
Mobile Communications
关键词
安全威胁
轻量级检测
流量指纹
蚁群算法
BP神经网络
security threat
lightweight detection
traffic fingerprint
ant colony algorithm
BP neural network
