Locomotion-based dynamic power management in embedded real-time systems

Most contemporary real-time distributed systems minimize computation energy consumption to reduce total system energy. However, wireless sensor networks expend a significant portion of total energy consumption on communication energy costs. Wireless transmission energy depends directly on the desired transmission distance, so the energy required for communication between neighboring nodes is less than that for distant ones. Mobile nodes can therefore reduce transmission energy costs by approaching one another before communicating. The penalty for energy reduction through locomotion is an increase in time consumed, thus care must be taken to meet system deadlines. We combine locomotion as a communication energy reduction strategy with well-known computation energy reduction schemes and demonstrate the resultant energy savings for representative systems.