wireless sensor network with energy hungry sensors

[attachment=18237]

INTRODUCTION
A wireless sensor network (WSN) consists of a large number of tiny sensor nodes deployed over a geographical area also referred as sensing field; each node is a low-power device that integrates computing, wireless communication and sensing abilities[1][2][3]. Nodes organize themselves in clusters and networks and cooperate to perform anassigned monitoring (and/or control) task without any human intervention at scales (both spatial and temporal) and resolutions that are difficult, if not impossible, to achieve with traditional techniques. Sensor nodes are thus able to sense physical environmental information (e.g., temperature, humidity, vibration, acceleration or whatever required), process locally the acquired data both at unit and cluster level, and send the outcome –or aggregated features- to the cluster and/or one or more collection points, named sinks or base stations



A General Framework for Energy-efficient Sensor Management

Most monitoring applications based on sensor networks rely on a synchronous philosophy where readings are carried out with a given sampling frequency. In such a case two main approaches can be considered to reduce the energy consumed by a sensor, i.e., duty cycling and adaptive sensing. Duty cycling consists in waking up the sensorial system only for the time needed to acquire a new set of samples and powering it off immediately afterwards. This strategy allows us for optimally managing energy provided that the dynamics of the phenomenon to be monitored are time-invariant and known in advance. Since such hypotheses only partly hold in many applications, periodic sensing is typically considered. Here, the (fixed) sampling rate is computed a priori, based on partial available information about the process to be monitored and assuming that the process dynamics are stationary. As a consequence, the sampling rate is larger than necessary (oversampling), e.g., 3 to 5 times, inducing, in turn, energy wasting. A better approach would require an adaptive sensing strategy able to dynamically adapt the sensor activity to the real dynamics of the process.


Hierarchical sensing
As mentioned above, hierarchical sensing techniques assume that multiple sensors are installed on the sensor nodes and observe the same phenomenon with a different resolution and power consumption (see Fig. 4). The idea behind hierarchical sensing techniques is to dynamically select which of the available sensors must be activated, by trading off accuracy for energy conservation.



Triggered sensing
The activation of the more accurate and power consuming sensors after the low-resolution ones once some activity within the sensed area has been detected is referred to triggered sensing. An example of triggered sensing is presented in [37] for structural health monitoring and damage detection of a civil structure (i.e., a bridge). The structure to be monitored is split into zones instrumented with sensing units capable of detecting two scales of responses: accelerometers (MEMS and piezo-electric) and strain gauges (the three-wire quarter-bridge circuit). A central node, which supervises all the activities of the sensor network, is endowed with a triggering system: sensor units are activated when the passage of isolated, large payload vehicles are detected by an imaging system [38]. Initially, in each sensor unit, only accelerometers are activated to collect data and perform a local assessment of the potential damage.