Outlier Cleaning and Sensor Data Aggregation Using Modified Z-score Technique

The outlier detection is an important preprocessing routine that is required to ensure robustness of the sensory data analysis. Several factors make the wireless sensor networks (WSNs) especially prone to outliers. A sensor network is equipped with thousands of inexpensive, low fidelity nodes, which can easily generate sensing errors. Since these networks include a large number of sensors, the chance of error accumulates. Besides, in their usage for security and military purposes, sensor nodes are susceptible to manipulation by adversaries. Hence, it is clear that outlier detection should be an inseparable part of any data processing routine that takes place in WSNs.