Delay Optimal Scheduling for Energy Harvesting Based Communications

Energy harvesting has attached considerably attention for its wide application in wireless green communications. Equipped with a rechargeable battery, a source node can harvest energy from ambient environments and rely on this free and regenerative energy supply to transmit packets. Due to the uncertainty of available energy from harvesting, however, it could induce intolerably large latency and packet loss, if the source always waits for harvest energy. To overcome this, the source would naturally resort to the reliable power grid for a quick delivery of backlogged packets. Thus, there exists a tradeoff between the average waiting time and power consumption on the reliable supply (which may require payment). In this paper, we attempt to address the efficient energy scheduling problem for a bursty communication link powered by a capacity-limited battery storing harvested energy and the grid. The proposed scheduling scheme would give priority to the usage of harvest energy and resort to the grid based on the current status of the queueing system. By a two-dimensional Markov chain modeling, we give an analysis on the average queuing delay and the average power consumption from the grid. Then, a linear programming problem is formulated to minimize the average delay under the constraint of a maximum allowable power consumption. By analyzing the corresponding optimization problem, we derive the optimal transmission policy and succeed to find the optimal transmission parameters. Our theoretical analysis is corroborated by simulation results.