Anomaly detection in wireless sensor network of the “smart home” system

Citations Over TimeTop 10% of 2017 papers

Abstract

Subject. The paper reviews the problem of anomaly detection in home automation systems. Authors define specificities of the existing security networks and accentuate the need of the detection of informational and physical impact on sensors. Characteristics of the transmitted information and physical impacts on automation devices are analysed and used as metrics for the anomalous behavior detection. Various machine learning algorithms for anomaly detection are compared and reviewed. Methods. The paper reviews the anomaly detection method that includes artificial neural networks as a detection tool. In this method characteristics of the security network devices are analysed to detect an anomalous behaviour, and exactly this type of data should be used to train the artificial neural network. This paper describes tools that can be used to implement the offered anomaly detection method. Main results. As an experiment the scenario has been created so that the model of the “Smart home” system produces the data of network information streams and the artificial neural network decides from this data. As a result the training and testing sets has been produced. The configuration of the artificial neural network has been defined as a result of tests. The experiment shows the potential of described method due to the fact that the area under ROC curve is 0.9689, which is better than basic machine learning algorithms performance. Practical importance. The offered method can be used at the development stage while implementation of the information and security systems requiring monitoring of the connected devices. Anomaly detection technology excludes the possibility of the inconspicuous violation of the information's confidentiality and integrity.

Related Papers

• → Explainable Anomaly Detection Framework for Maritime Main Engine Sensor Data(2021)59 cited
• → Towards Experienced Anomaly Detector Through Reinforcement Learning(2018)56 cited
• → Anomaly Detection with Partially Observed Anomaly Types(2021)4 cited
• → Human-machine interactive streaming anomaly detection by online self-adaptive forest(2022)8 cited
• → Tree-based Self-adaptive Anomaly Detection by Human-Machine Interaction(2021)1 cited