Sensing uncertainty reduction using low complexity actuation

  • Aman Kansal ,
  • Eric Yuen ,
  • William J. Kaiser ,
  • Gregory J. Pottie ,
  • Mani B. Srivastava

ACM Symposium on Information Processing in Sensor Networks (IPSN) |

Published by ACM Press

The performance of a sensor network may be best judged by the quality of application specific information return. The actual sensing performance of a deployed sensor network depends on several factors which cannot be accounted at design time, such as environmental obstacles to sensing. We propose the use of mobility to overcome the effect of unpredictable environmental influence and to adapt to run time dynamics. Now, mobility with its dependencies such as precise localization and navigation is expensive in terms of hardware resources and energy constraints, and may not be feasible in compact, densely deployed and widespread sensor nodes. We present a method based on low complexity and low energy actuation primitives which are feasible for implementation in sensor networks. We prove how these primitives improve the detection capabilities with theoretical analysis, extensive simulations and real world experiments. The significant coverage advantage recurrent in our investigation justifies our own and other parallel ongoing work in the implementation and refinement of self-actuated systems.