Age-Optimal Joint Sampling and Transmitting Scheduling for Wireless Sensor Networks with Energy Harvesting

This paper considers the age-optimal joint sampling and transmitting scheduling design for wireless sensor networks with energy harvesting. The optimization problem is modeled as a Markov decision process (MDP), and a low-complexity online policy is proposed based on the formulation, which has two main steps, i.e., “separate deciding” and “joint scheduling”. Specifically, each sensor's action is determined by the double-threshold policy first and then the scheduled sensor is chosen from the set of non-idle sensors by comparing the expected reward gain. The proposed policy can be extended to the case of partial state information knowledge, where the state information at the sensors' side can only be delivered through packet updatings. Numerical results show that our proposed policy outperforms the existing benchmarks under general settings.