Due to the limited power supply and available bandwidth of sensor nodes, 
ensuring reliability and long lifetime of a sensor network requires the 
adoption of adequate strategies to schedule the participation of the nodes in 
sensing and communication. Beyond the role that medium access control and 
routing protocols may play, the application logic can support such 
optimization by selecting subsets of sensor nodes whose readings alone can 
achieve the application-dependent quality standards. In this paper, we focus on 
random sensor selection strategies for wireless sensor networks and 
investigate the potential for optimizing the sensing scheduling of the nodes 
in the spatial dimension. In particular, we propose ARS, an adaptive random 
sensor selection scheme that relies upon the CTP data collection protocol to 
collect information about the neighborhood of each node within a region of 
interest. ARS then uses this information to determine, in a distributed 
manner, the probability with which each node should participate in sensing and 
data communication. We support our investigations through a preliminary 
simulation study and demonstrate the ability of ARS to provide for good 
sensing coverage of the area of interest while limiting the number of sensing 
nodes and, thus, the amount of data communication.