TD3-Based Resource Allocation for Low-Power AmBC-Assisted WBANs in Cognitive Radio Networks

The Internet of Medical Things (IoMT) provides a promising paradigm for remote health care. However, energy limitations and spectrum shortage issues limit the health monitoring applications of wearable sensors in wireless body area networks (WBANs). Since cognitive radio networks (CRNs) and ambient backscatter communication (AmBC) techniques allow spectrum sharing and energy harvesting from ambient radio frequency (RF) signals for wearable sensors, there are potential benefits of combining CRN and AmBC with WBAN to support sustainable health monitoring. In this paper, we propose a low-power AmBC-assisted WBAN health monitoring system model in a CRN, where the wearable sensor nodes are able to harvest energy to maintain operation and transmit signals while the primary transmitter (PT) is idle. By considering the unique characteristics of WBAN channels, different data rate and energy requirements of different types of wearable sensors, and random data collection, we formulate the joint resource and transmission management (JRTM) optimization problem including data queue backlog minimization, transmission rate maximization, and energy management. We also propose a twin delayed deep deterministic policy gradient (TD3) based intelligent joint resource allocation (IJRA) algorithm to solve the proposed stochastic optimization problem and maintain the sustainability of the proposed system without prior knowledge of the environment. Extensive numerical simulation results demonstrate that our proposed algorithm can effectively control the data queue backlog under random data collection, and represents better stability, adaptability, and sustainability compared to five benchmark policies to support personalized health monitoring.