Learning Aided Joint Sensor Activation and Mobile Charging Vehicle Scheduling for Energy-Efficient WRSN-Based Industrial IoT

Actor-critic method for the MCV charging scheduling.

Abstract

In this paper, the joint sensor activation and mobile charging vehicle scheduling for wireless rechargeable sensor network (WRSN) based industrial Internet of Things (IIoT) is studied. In the proposed framework, an optimal sensor set is selected to collaboratively execute a bundle of heterogeneous industrial tasks (e.g., production-line monitoring), meeting the quality-of-monitoring (QoM) of each individual task, and we consider that a mobile charging vehicle (MCV) is scheduled for recharging sensors before their charging deadlines, i.e., time instants of running out of their batteries, in order to prevent from any potential service interruptions (which is one of the key features of IIoT). Our goal is to jointly optimize the sensor activation and MCV charging scheduling for minimizing the system energy consumption, subject to tasks’ QoM requirements, sensor charging deadlines and the energy capacity of the MCV. Unfortunately, solving this problem is nontrivial, because it involves solving two tightly coupled NP-hard optimization problems. To address this issue, we design a novel scheme integrating reinforcement learning and marginal product based approximation algorithms, and prove that it is not only computationally efficient but also theoretically bounded with a guaranteed performance in terms of the approximation ratio. Simulation results show the feasibility of the proposed scheme and demonstrate its superiority over counterparts.

Publication
IEEE Transactions on Vehicular Technology (IF: 6.8, Q1)
If you want cite this work, please click the Cite button above to export the bib format. Thank you! 😊
Jiayuan Chen
Jiayuan Chen
PhD Candidate in Computer Science and Technology

My research interests include Human Digital Twin (HDT), Network Resource Management, Edge Computing and Edge Intelligence, Tactile Internet (TI), and Data-Driven Optimization and Learning.