ABSTRACT

In Wireless Sensor Network (WSN), considerable number of smart sensor nodes are interconnected to carry out a specific set of sensing task within the geographic area. WSN is generally heavily resource constrained. Hence judicious placement of optimal number of sensor nodes is of high importance to produce meaningful inference from the sensed data. Improper and unplanned sensor placement for event sensing produces trivial data and that too at the cost of higher network resource utilization. It reduces efficiency of the overall network. This work proposes a wiser criterion for sensor node selection that would produce optimal information utility from the sensed data. The proposed criterion is based on the performance of specific sensor-positioning approaches where the performance measure is the accuracy of estimated location information produced by each of them. The proposed approaches provide a quantitative basis for communication-link selection during network discovery and self-localization. The same optimal criteria may also provide theoretical basis for node-selection towards achieving optimum accuracy in target localization by the sensor network.