A Novel Ensemble Approach for Anomaly Detection in Wireless Sensor Networks Using Time-overlapped Sliding Windows

One of the most important issues concerning the sensor data in the Wireless Sensor Networks (WSNs) is the unexpected data which are acquired from the sensors. Today, there are numerous approaches for detecting anomalies in the WSNs, most of which are based on machine learning methods. In this research, we present a heuristic method based on the concept of “ensemble of classifiers” of data mining.  Our proposed algorithm, at first applies a fuzzy clustering approach using the well-known C-means clustering method to create the clusters. In the classification step, we created some base classifiers, each of which utilizes the data of overlapping windows to utilize the correlation among data over time by creating time-overlapped batches of data. By aggregating these batches, the classifier proceeds to find an appropriate label for future incoming instance. The concept of “Ensemble of Classifiers” with majority voting scheme has been used in order to combine the judgment of all classifiers. The results of our implementation with MATLAB toolboxes shows that the proposed majority-based ensemble learning method attains more efficiency compared to the case of the single classifier method. Our proposed method enhances the performance of the system in terms of major criteria such as False Positive Rate, True Positive Rate, False Negative Rate, True Negative Rate, Sensitivity, Specificity and also the ROC curve.