Spatio-Temporal Correlation based Anomaly Detection and Identification Method for IoT Sensors

Status monitor and anomaly detection of sensor nodes is critical for reliability of internet of things (IoT) system. Due to the complex causes of sensor anomalies, traditional methods that estimate only sensor state or environment state when detect anomalies cannot satisfy practical requirements. And existing methods often neglect spatial and temporal dependencies among IoT sensor observations. This paper defines faulty nodes and event nodes and proposes a novel anomaly detection and identification method implemented in two stages: 1) Anomaly detection stage by a customized composite distance metric and sensor clustering. 2) Anomaly source identification stage by fuzzy logic system based on spatio-temporal correlation. Experiments demonstrated that the proposed method can effectively detect anomaly sensors and capture spatio-temporal correlation to identify the anomaly source.

[1]  Harsh K. Verma,et al.  Outlier Detection of Data in Wireless Sensor Networks Using Kernel Density Estimation , 2010 .

[2]  Pabitra Mohan Khilar,et al.  Fault Diagnosis in Wireless Sensor Networks: A Survey , 2013, IEEE Communications Surveys & Tutorials.

[3]  Qiu Tong.  A method for sensor fault diagnosis based on wavelet transform and neural network , 2013 .

[4]  Vallipuram Muthukkumarasamy,et al.  Flow-Based Anomaly Detection Using Neural Network Optimized with GSA Algorithm , 2013, 2013 IEEE 33rd International Conference on Distributed Computing Systems Workshops.

[5]  S. Hashemi,et al.  I-IncLOF: Improved incremental local outlier detection for data streams , 2012, The 16th CSI International Symposium on Artificial Intelligence and Signal Processing (AISP 2012).

[6]  Y. Zhang,et al.  – 20 Statistics-based outlier detection for wireless sensor networks , 2012 .

[7]  Zhanxing Zhu,et al.  Spatio-temporal Graph Convolutional Neural Network: A Deep Learning Framework for Traffic Forecasting , 2017, IJCAI.

[8]  Christopher Leckie,et al.  High-dimensional and large-scale anomaly detection using a linear one-class SVM with deep learning , 2016, Pattern Recognit..

[9]  Weili Wu,et al.  Localized Outlying and Boundary Data Detection in Sensor Networks , 2007, IEEE Transactions on Knowledge and Data Engineering.

[10]  Manju Bala,et al.  Fault Diagnosis in Wireless Sensor Networks - A Survey , 2018, 2018 4th International Conference on Computing Sciences (ICCS).

[11]  Wang Hong,et al.  Effective method of sensor fault diagnosis , 2012 .

[12]  Jianzhong Li,et al.  Unsupervised Outlier Detection in Sensor Networks Using Aggregation Tree , 2007, ADMA.

[13]  Matej Mohar,et al.  Security in Internet of Things , 2017 .

[14]  Eyal Amir,et al.  Real-time Bayesian Anomaly Detection for Environmental Sensor Data , 2007 .

[15]  Jiafu Wan,et al.  Security in the Internet of Things: A Review , 2012, 2012 International Conference on Computer Science and Electronics Engineering.

[16]  Hans-Peter Kriegel,et al.  LOF: identifying density-based local outliers , 2000, SIGMOD '00.