Detección de anomalías en redes de sensores inalámbricos

Abstract

Over time, wireless sensor networks (WSN) have been used for a variety of applications. Extensive work has been dedicated to various WSN applications. It is important to note that, due to their physical limitations, the sensors are prone to several types of faults. These restrictions can pose serious problems in event detection applications. Especially if the WSNs are deployed in hostile environments, such as the industrial or environmental sector. The detection of anomalies has recently attracted the attention of the scientific community, due to its relevance in real-world applications. The proposed solutions depend to a large extent on supervision and communication, using techniques based on tools such as Machine Learning and Neural Networks. In this context, we introduce the most commonly used anomaly detection techniques in WSN. Compiling and comparing the main methods applied in specific scenarios, we analyze the advantages and conveniences of using any of them.