Adaptive compressive sensing based sample scheduling mechanism for wireless sensor networks

Sample scheduling is a crucial issue in wireless sensor networks (WSNs). The design objectives of efficient sample scheduling are in general two-folds: to achieve a low sample rate and also high sensing quality. Recently, compressive sensing (CS) has been regarded as an effective paradigm for achieving high sensing quality at a low sample rate. However, most existing work in the area of CS for WSNs use fixed sample rates, which may make sensor nodes in a WSN unable to capture significant changes of target phenomenon, unless the sample rate is sufficiently high, and thus degrades the sensing quality. In this paper, to pursue high sensing quality at low sample rate, we propose an adaptive CS based sample scheduling mechanism (ACS) for WSNs. ACS estimates the minimum required sample rate subject to given sensing quality on a per-sampling-window basis and accordingly adjusts sensors' sample rates. ACS can be useful in many applications such as environment monitoring, and spectrum sensing in cognitive sensor networks. Extensive trace-driven experiments are conducted and the numerical results show that ACS can obtain high sensing quality at low sample rate.