Anomaly Detection using Self-Organizing Map and Wavelets in Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have been developed and extensively applied in agriculture monitoring to monitor and collect various physical attributes within a specific area or environment of interest. Data readings from the sensors may be abnormal due to the sensors themselves such as limited battery power, onboard processing capability, sensor malfunction, or noise from the communication channel. It is thus, important to detect such data anomalies available in WSNs to determine a suitable course of action. The underlying aim of this paper is therefore to propose an anomaly detection scheme which is able to detect anomalies accurately by means of exploiting both time and frequency characteristics of the data signals. The contribution of this paper centers on anomaly detection by using Discrete Wavelet Transform (DWT) combined with a competitive learning neural network called self-organizing map (SOM) in order to accurately detect abnormal data readings are collected from WSNs.