Anomaly Detection Using Dynamic Sliding Window in Wireless Body Area Networks

Abstract

Anomaly detection is one of the critical challenges in Wireless Body Area Networks (WBANs). Faulty measurements in applications like health care lead to high false alarm rates in the system which may sometimes even causes danger to human life. The main motivation of this paper is to decrease false alarms thereby increasing the reliability of the system. In this paper, we propose a method for detecting anomalous measurements for improving the reliability of the system. This paper utilizes dynamic sliding window instead of static sliding window and Weighted Moving Average (WMA) for prediction purposes. The propose method compares the difference between predicted value and actual sensor value with a varying threshold. If average of the number of parameters exceed the threshold, true alarm is raised. Finally we evaluate the performance of the proposed model using a publicly available dataset and has been compared with existing approaches. The accuracy of the proposed system is evaluated with statistical metrics.