A LED light control system based on user's location using battery-less human detection sensor networks is designed and implemented in an indoor office environment to satisfy user's luminance requirement and reduce office's lighting energy consumption. In order to achieve high energy stability, easy maintenance and flexible deployment, all sensors are battery-less and activated by wireless energy transmitters embedded in LED ceiling lights, so that sensors can be located freely in the room. The system employs low energy-cost binary infrared human detection sensors, which however has large false detection errors. Therefore we deploy a distributed set of sensors, and user's motion is detected, tracked and predicted by a maximum likelihood algorithm only using sensing data from the multiple sensors. LED lights' switch states are controlled based on the likeli-hood map and required luminance. Experiment results confirm that our light control system can reduce the energy consumption significantly and satisfy user's luminance requirement with 100% probability.
