Anomalies Detection in Wireless Sensor Networks Using Bayesian Changepoints

Wireless Sensor Networks (WSN) have gained wide acceptance as a framework for telemetry and remote monitoring for applications such as telemedicine, precision agriculture, and climate monitoring. The complex and dynamic characteristics of such networks have made them vulnerable to anomalies, i.e., observations that do not correspond to the natural behavior of measurements. This paper evaluates the use of Bayesian Changepoints in the context of anomalies detection in WSNs, in order to determine under which conditions this technique leads to the minimization of false positives.