Generic correlation model for wireless sensor network applications

In wireless sensor networks (WSNs) having a high density of sensor nodes, transmitted measurements are spatially correlated, often redundantly whenever an event of interest is detected. In this work, the authors propose a correlation model to enable energy-efficient methodologies that exploit the spatial correlation at the network and medium access control (MAC) layers. At the network layer, the authors first demonstrate how, through proper tuning of both the sensing range and correlation threshold, WSNs can be partitioned into disjoint correlated clusters without degrading the information reliability, thus enabling significant energy saving. On the other hand, as another contribution, the authors investigate the impacts of correlation between nodes on achieved distortion in the event estimation at the sink. The interactions among various parameters such as distortion constraints, spatial node density, node selection, sensing range and their impacts on the reconstruction performance are quantitatively studied. The authors demonstrate that the same level of distortion constraint can be achieved by selecting fewer nodes. The nodes can be still fewer if they are less spatially correlated. However, the MAC protocols which show distributive effect of this selection needs more study.