Control-Oriented Transmission Power Optimization for NOMA-Based Multi-Loop WNCSs

The recent advent of artificial intelligence and 6G technologies has catalyzed a novel research trend towards goal-oriented design for multi-user remote control systems. Existing works have rarely simultaneously focused on the direct characterization for effective control, and targeted schemes for multiuser scenarios, e.g., non-orthogonal multiple access (NOMA). In this paper, we propose a control-oriented NOMA system capable of intelligent transmission power control. Through theoretical derivation, we build up an age of information (AoI)-dependent function of control cost within the NOMA system framework. In pursuit of the optimal trade-off between control cost and power consumption, we formulate an optimization problem as a Markov decision process and employ the Dueling-Double-Deep Q Network (D3QN) to intelligently decide the transmission power. Numerical results demonstrate the superiority of the control-oriented NOMA system over orthogonal multiple access (OMA) in achieving high-quality multi-loop control while considering energy expenditure.