Interference management in NOMA-enabled virtualized wireless networks

• Published in: Wireless networks Vol. 28; no. 4; pp. 1457 - 1474
• Main Authors: Liu, Chengyi, Tao, Yu, Xing, Song
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.05.2022; Springer Nature B.V
• Subjects: Algorithms; Communications Engineering; Complexity; Computer Communication Networks; Computing costs; Electrical Engineering; Engineering; Interference; IT in Business; Networks; Nonorthogonal multiple access; Original Paper; User requirements; Virtual networks; Wireless networks; Interference management; Virtualized wireless networks; Non-orthogonal multiple access; Resource allocation
• ISSN: 1022-0038; 1572-8196
• DOI: 10.1007/s11276-022-02911-3

In this paper, we address the interference management problem in non-orthogonal multiple access (NOMA)-enabled virtualized wireless networks (VWNs) by using power allocation approaches. Specifically, the power resources of the base station (BS) are shared among different service providers (called the slices), where the maximum tolerant interference is considered for each slice to guarantee their interference isolation. The interference management (IM) problem is formulated aiming to maximize the sum-rate of the system subject to the slice interference isolation, the minimum required rates of the individual users, and the power budget constraints of the whole system. Then, an optimal interference management algorithm (IMA) is proposed to solve the IM problem in a centralized manner at the BS. In addition, a computational-complexity reduced IMA (CCRIMA) is proposed with the implementation in a semi-distributed manner within each slice to obtain a suboptimal IM solution. Simulation results show that the proposed optimal power allocation in IMA achieves a flexible interference management within each slice, while supporting the minimum rate requirements of all users by adjusting the maximum tolerant interference in the slice. Moreover, the proposed CCRIMA can approximate the optimal performance of IMA in terms of the sum-rate of the system with a little computational cost of the computational complexity in each slice.