Optimal scheduling for energy harvesting mobile sensing devices

The rapid advances in mobile devices and their embedded sensors have enabled a compelling paradigm for collecting ubiquitous data to share with each other or the general public. In this paper, we study how to achieve the close-to-optimal transmission utility performance for sensor-enhanced mobile devices that are capable of harvesting energy from the environment. This is a very challenging task due to the stochastic and unpredictable nature of data arrival, channel condition, and energy replenishment. By taking advantage of the Lyapunov optimization framework, we propose an online scheduling algorithm called OSCAR (Optimal SCheduling AlgoRithm), which jointly make control decisions on system state, energy harvesting, and data transmission for achieving optimal utility on mobile sensing devices. Different from traditional techniques, OSCAR does not require any knowledge of system statistics, including the energy state process. Rigorous analysis and extensive experiments have demonstrated both the system stability and the utility optimality achieved by the OSCAR algorithm.