Wake-up Control for Energy-efficient Identifications of Multiple Emission Sources in Wireless Sensor Networks.

This paper investigates the energy–efficient wake–up control to identify the locations of multiple sources emitting a sensing observation, e.g., toxic gas, which are spatially spread over a sensing area in wireless sensor networks. In order to identify multiple sources efficiently in terms of energy consumed by sensor nodes, it is desirable to activate only a subset of nodes contributing to the increase of the accuracy of identification. To this end, in this paper, we investigate how to estimate the location of multiple emission sources of sensing observation by using the collected subset of data owning different attributes. As wake–up control to activate sensor nodes, we assume two types of wake–up control: unicast wake–up (UCWu), which individually activates nodes based on ID allocated to each node, and content–based wake–up (CoWu), which can realize activations of nodes based on their sensing data. We evaluate the identification accuracy of different approaches by computer simulations, and clarify wake–up control suited for energy–efficient identification of multiple emission sources.