LAUSR

In this article, an energy-efficient temporal sensing framework is proposed for a wireless sensor node (SN) monitoring a temporal process. A concept of Age of Sample (AoS) is introduced, and expressions for the AoS and average AoS functions are derived that capture freshness of sensed samples (or intersample time). To incorporate the effect of process variations, weighted AoS (WAoS) and its average functions are developed. The framework solves a multiobjective optimization problem (MOP) that optimizes this average WAoS function and energy efficiency of the SN to select a few temporal sensing instances of a sensing window while maintaining a predefined sensing quality. An upper bound on the average WAoS is derived, which makes the MOP solvable. Using these few measurements, the process signal across entire sensing window is estimated by leveraging inherent temporal correlation. The idea of adapting sensing window according to the changing correlation of the process is also presented, which addresses the nonstationary aspect of the monitored process. Simulation studies on real data sets illustrate that on comparison with the closest existing scheme, the proposed scheme provides 30.1% gain in sensing quality while consuming nearly same sensing energy. Furthermore, it consumes 22.4% lesser sensing energy to maintain nearly same sensing quality.