Minimizing the AoI for Pull-Based Target-Level Data Collection in Energy-Harvesting IoTs

Data collection is a crucial task of IoTs. According to the data collection scheme and the required data granularity, data collection in IoTs can be classified into pull-based/push-based data collection as well as node-level/target-level data collection. Thus, there are four scenarios for data collection: Push-based node-level data collection (Push-Node), Push-based target-level data collection (Push-Target), Pull-based node-level data collection (Pull-Node), and Pull-based target-level data collection (Pull-Target). Energy-Harvesting IoT (EH-IoT) is an important component of IoTs and the Age of Information (AoI) minimization problem has been studied extensively for data collection in EH-IoTs. However, existing works only studied the problem under the scenario of Push-Node, Push-Target and Pull-Node. Therefore, this paper investigates the AoI minimization problem for Pull-based Target-level data collection in EH-IoTs (AoI-Pull-Target) for the first time. AoI-Pull-Target is formally defined and proved to be NP-hard. A two-stage dynamic programming-based node scheduling algorithm and a real-time schedule adjustment scheme are proposed to solve the problem. The proposed algorithm is analyzed theoretically. Extensive simulations and real-world testbed experiments verify the high performance of our algorithm.