| Hi-index | 0.00 |
Nowadays, wireless sensor networks are being used for a fast-growing number of different application fields (e.g., habitat monitoring, highway traffic monitoring, remote surveillance). Monitoring (i.e., querying) the sensor network entails the frequent acquisition of measurements from all sensors. Since sensor data acquisition and communication are the main sources of power consumption and sensors are battery-powered, an important issue in this context is energy saving during data collection. Hence, the challenge is to extend sensor lifetime by reducing communication cost and computation energy. This paper thoroughly describes the complete design, implementation and validation of the SeReNe framework. Given historical sensor readings, SeReNe discovers energy-saving models to efficiently acquire sensor network data. SeReNe exploits different clustering algorithms to discover spatial and temporal correlations which allow the identification of sets of correlated sensors and sensor data streams. Given clusters of correlated sensors, a subset of representative sensors is selected. Rather than directly querying all network nodes, only the representative sensors are queried by reducing the communication, computation and power costs. Experiments performed on both a real sensor network deployed at the Politecnico di Torino labs and a publicly available dataset from Intel Berkeley Research lab demonstrate the adaptability and the effectiveness of the SeReNe framework in providing energy-saving sensor network models.