ANN-Based outlier detection for wireless sensor networks in smart buildings

Wireless sensor network (WSN) is an emerging technology with a wide range of potential applications in smart buildings. The measuring process by using WSNs in the actual environment always introduces noise, errors, accidents, and other potential outliers to the data collected by the sensors. It is crucial to establish an effective approach for outlier detection and recovery in the real applications of WSNs. In this paper, we propose an outlier detection and recovery approach using artificial neural network (ANN), which can be used to determine whether the temperature values measured by the sensors in WSNs are outliers. The experimental results in real building show that the proposed ANN-based models can provide a reasonably good prediction of the temperature and high accuracy in buildings compared with the hidden Markov model (HMM)-based approach, which can potentially be used for outlier detecting and thermal controlling in the Internet of Things (IoT) applications.