Wireless Sensor Network Design for Energy-Efficient Monitoring

Application Scenarios

Some contexts which would benefit from the proposed approach are described in the following. (1) Agricultural production monitoring (Burrell at al., 2004). A sensor network can be exploited in agricultural production (1) to identify the risk of frost damage to vines, (2) to assess the risk of powdery-mildew outbreak (or to detect pests and irrigation needs), or (3) to detect the presence of birds. A trial sensor network (involved 18 nodes) has been deployed in a local Oregon vineyard to collect different measures (e.g., temperature, lighting levels, humidity, presence of birds) for several weeks during the summer of 2002 (Burrell at al., 2004). By means of this deployment it has been possible to observe some correlations among sensor data. There is a great variability across the vineyard during the day and less variation during the night, hence measurements are more correlated during the night and less during the day. Furthermore, there are different seasonal issues (e.g., risk of frost damage to vines). For example, during the winter a wireless sensor network can be exploited in an agricultural production to gather frequent temperature data and to alert the system only when a risk of frost damage is detected (i.e., temperature is lower than a given threshold). Hence, sensor data are correlated both in time and space, and a more power-efficient technique would be necessary to efficiently collect the required information, optimize sensor placement and extend the sensor network lifetime.