Optimal Adaptive Modulation for QoS Constrained Wireless Networks with Renewable Energy Sources

In this paper we derive the optimal policy to minimize the average number of dropped packets for a delay constrained wireless node with a renewable energy source. The proposed framework employs adaptive modulation for transmission of the optimal number of packets towards satisfying the Quality of Service (QoS) constraints. This framework is formulated as a Markov Decision Process (MDP) which minimizes the long term packet drop rate. Further, we demonstrate that the optimal policy is monotone with respect to each of the joint state components of the MDP. Simulation results are presented to validate the monotone structure and superior performance of the proposed policy with respect to the existing schemes.