SILENCE: distributed adaptive sampling for sensor-based autonomic systems

  • Authors:
  • Eun Kyung Lee;Hariharasudhan Viswanathan;Dario Pompili

  • Affiliations:
  • Rutgers University/ NSF Center for Autonomic Computing, New Brunswick, NJ, USA;Rutgers University/ NSF Center for Autonomic Computing, New Brunswick, NJ, USA;Rutgers University/ NSF Center for Autonomic Computing, New Brunswick, NJ, USA

  • Venue:
  • Proceedings of the 8th ACM international conference on Autonomic computing
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Adaptive sampling and sleep scheduling can help realize the much needed resource efficiency in densely deployed autonomic sensor-based systems that monitor and reconstruct physical or environmental phenomena. This paper presents a data-centric approach to distributed adaptive sampling aimed at minimizing the communication and processing overhead in autonomic networked sensor-based systems. The proposed solution exploits the spatio-temporal correlation in sensed data and eliminates redundancy in transmitted data through selective representation without compromising on accuracy of reconstruction of the monitored phenomenon at a remote monitor node. In addition, the solution also exploits the same correlations for adaptive sleep scheduling aimed at saving energy in Wireless Sensor Networks (WSNs) while also providing a mechanism for ensuring connectivity to the monitor node. The data-centric joint adaptive-sampling and sleep-scheduling solution, SILENCE, has been evaluated through real experiments on a testbed monitoring temperature and humidity distribution in a rack of servers as well as through extensive simulations on TOSSIM, the TinyOS simulator.