In this paper, we propose a model-based approach to detect outliers insensor networks by exploring the spatial correlation among neighboringnodes. This research is motivated by the observation that sensors in closeproximity normally present similar readings.We propose to employ Gaussianmixture modeling as a statistical means to build a probability densityfunction for multivariate spatial neighborhood sensor readings. Outlyingsensors can be reliably detected since they exhibit extremely low densityvalues. Our extensive simulation evaluation validates the proposedmodel-based approach for outlier detection in sensor networks.
|Number of pages||19|
|Journal||Ad-Hoc and Sensor Wireless Networks|
|State||Published - 2011|
- Expectation-maximimation;wireless sensor networks
- Gaussian mixture model
- Outlier detection