This paper was presented at the SystemSim workshop in the International Conference on Wireless and Mobile Communications, held on November 13-17, 2016 in Barcelona, Spain.
Optimization of the energy consumption of sensor networks is traditionally performed either at the sensor node level or at the network level. However, more energy savings can be obtained if the application that makes use of the sensor nodes is considered. In order to achieve such extra energy gains, control theory can be applied. This paper summarizes control strategies implemented to minimize the energy consumption of a set of sensor nodes, while ensuring the application Quality of Service, this latter being mainly expressed with a minimum number of samples that must be available at the application level at each sampling time. With the Control strategies proposed, the sensor network lifetime is increased compared to the case without control strategy at application level. The control strategies have been implemented and evaluated on a real testbed composed of heterogeneous sensor nodes, and using the LINC middleware for node coordination.