학술논문

Energy-Efficient THz NOMA-Enabled Small Cells Underlaying Macrocell Using Reinforcement Learning
Document Type
Conference
Source
2023 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS) Advanced Networks and Telecommunications Systems (ANTS), 2023 IEEE International Conference on. :1-6 Dec, 2023
Subject
Communication, Networking and Broadcast Technologies
Computing and Processing
Signal Processing and Analysis
Wireless communication
Training
NOMA
Power control
Reinforcement learning
Downlink
Energy efficiency
DRL
DDPG
EE
THz
6G
Language
ISSN
2153-1684
Abstract
The Terahertz (THz) frequency range has attracted significant interest owing to its exceptional high frequency and broad bandwidth that can be easily accessed. THz technology is a crucial component of the sixth generation (6G) wireless communication networks. In this research, we have incorporated the downlink non-orthogonal multiple access (NOMA) technology in small cell (SC) networks operating in the THz band to enhance the overall performance of the network. Despite the above mentioned advantage, the combination of these technologies increase the energy consumption. So, to address this problem, we formulated a problem to maximize the energy efficiency (EE) of THz-NOMA downlink network by optimizing resource block (RB) assignment and power control. To achieve the target, we used deep deterministic policy gradient (DDPG) technique because it has an ability to solve the continuous action spaces as compared to traditional model-based approaches. Numerical results demonstrated that the proposed scheme acquire better results than the state-of-art schemes.