SILENCE: distributed adaptive sampling for sensor-based autonomic systems
Proceedings of the 8th ACM international conference on Autonomic computing
Wireless Personal Communications: An International Journal
| Hi-index | 0.00 |
Data samples sensed from a densely deployed sensor network often exhibit strong spatial and temporal correlation. Sampling and transmitting these highly correlated data through wireless channel would result in excessive energy consumption. One way to conserve energy is to set a portion of sensors into sleep mode provided their missing data can be well recovered by sensors that remain active. Different from traditional full area coverage sleep scheduling schemes, we propose a data-coverage scheme based on the ability of active nodes to recover the missing data of sleeping nodes. For this purpose, we present the scheme to extract correlation among sensing data of nodes. We also propose the Greedy Reduction Algorithm (GRA) to select the set of active sensors. The missing data of sleeping nodes is recovered using active nodes’ sensing data as well as the correlation among data of nodes. Using a publicly released sensing data set, we analyze how to implement this data-coverage sleep scheduling scheme. We also evaluate its performance and conduct comparison experiments with the full area coverage scheme. Experimental results indicate that the data-coverage sleep scheduling scheme offers superior performance.