Collaborative processing in Mote-based sensor/actuator networks for environment control application

This paper describes in the framework of optimization a collaborative sensing and actuation system for environment control. In the collaborative sensing, the sensor network topology is self-configured according to the sensing information to optimize a sensing utility. Experimental results using Motes show that the algorithm provides the sensor network topology allocating resources to sensing nodes. The server can robustly gather the sensing data from all sensor nodes using the collaborative sensing algorithm, and calculate the control signals for actuators to balance the energy saving against the quality of the control signals. In addition to a centralized approach, a decentralized algorithm is also proposed to calculate the control signals. Simulations reveal that the decentralized algorithm can provide the same performance as the centralized approach. We also demonstrate its accuracy and efficiency performance using the Motes and compare it with the simulation.