학술논문
Performance Analysis and Learning-Assisted Power Control for NOMA Enabled D2D-Cellular Network
Document Type
Periodical
Author
Source
IEEE Systems Journal Systems Journal, IEEE. 18(1):278-281 Mar, 2024
Subject
Language
ISSN
1932-8184
1937-9234
2373-7816
1937-9234
2373-7816
Abstract
This work investigates a device-to-device (D2D) underlayed cellular system where both D2D and cellular networks are NOMA enabled, which is not only more spectrally efficient than the previous D2D and NOMA models but also can outperform them. Specifically, we first present closed-form expressions for system outage probability (SOP) and sum ergodic rate (SER) metrics for performance analysis and thereafter propose a deep neural network-based power control mechanism for SOP minimization. Analytical results are validated with extensive simulations that reveal the effectiveness of the proposed model over comparative schemes and the requirement of optimizing the power values in accordance with change in different system parameters.