Power control and task offloading strategies for high-density wireless body area networks based on deep reinforcement learning

With the widespread application of Wireless Body Area Network (WBAN) in healthcare, issues such as a lack of computational resources and communication interference in high-density WBAN scenarios have become increasingly prominent. To address these issues, this paper introduces Mobile Edge Computing (MEC) to tackle the lack of computational resources and employs power control to mitigate communication interference among WBANs. We describe the joint optimization problem of power control and task offloading in high-density WBAN scenarios as a Markov Decision Process (MDP) and propose a Power Control and Task Offloading (PCTO) algorithm based on the Deep Deterministic Policy Gradient (DDPG) method to achieve coordinated optimization of transmission power and task offloading. Furthermore, we improve the algorithm’s learning efficiency and sample utilization by incorporating a prioritized experience replay mechanism. By constructing a multi-objective reward function that comprehensively considers delay, energy consumption, and Signal-to-Interference-plus-Noise Ratio (SINR), and incorporating Quality of Service (QoS) constraints, the algorithm is capable of making intelligent decisions in complex wireless communication environments. The simulation results demonstrate that, compared to existing methods, the proposed approach significantly reduces system delay while improving SINR and service satisfaction in various WBAN density scenarios.