[Paper Review] Resource Optimization and Power Allocation in In-band Full Duplex (IBFD)-Enabled Non-Orthogonal Multiple Access Networks

Mohammed S. Elbamby, Mehdi Bennis+3 more|arXiv (Cornell University)|Jun 17, 2017

Full-Duplex Wireless Communications4 citations

TL;DR

This paper proposes a joint user association, mode selection (HD/FD, OMA/NOMA), and power allocation framework for in-band full-duplex (IBFD) and non-orthogonal multiple access (NOMA) networks using Lyapunov optimization. It achieves up to 63% and 73% gains in UL and DL throughput, and 21% and 17% gains in cell-edge throughput, respectively, under dynamic interference and queue stability constraints.

ABSTRACT

In this paper, the problem of uplink (UL) and downlink (DL) resource optimization, mode selection and power allocation is studied for wireless cellular networks under the assumption of in-band full duplex (IBFD) base stations, non-orthogonal multiple access (NOMA) operation, and queue stability constraints. The problem is formulated as a network utility maximization problem for which a Lyapunov framework is used to decompose it into two disjoint subproblems of auxiliary variable selection and rate maximization. The latter is further decoupled into a user association and mode selection (UAMS) problem and a UL/DL power optimization (UDPO) problem that are solved concurrently. The UAMS problem is modeled as a many-to-one matching problem to associate users to small cell base stations (SBSs) and select transmission mode (half/full-duplex and orthogonal/non-orthogonal multiple access), and an algorithm is proposed to solve the problem converging to a pairwise stable matching. Subsequently, the UDPO problem is formulated as a sequence of convex problems and is solved using the concave-convex procedure. Simulation results demonstrate the effectiveness of the proposed scheme to allocate UL and DL power levels after dynamically selecting the operating mode and the served users, under different traffic intensity conditions, network density, and self-interference cancellation capability. The proposed scheme is shown to achieve up to 63% and 73% of gains in UL and DL packet throughput, and 21% and 17% in UL and DL cell edge throughput, respectively, compared to existing baseline schemes.

Motivation & Objective

• Address the challenge of interference and spectral efficiency in dense cellular networks with in-band full-duplex (IBFD) base stations and non-orthogonal multiple access (NOMA).
• Optimize uplink and downlink resource allocation, user association, and transmission mode selection (HD/FD, OMA/NOMA) under queue stability constraints.
• Maximize time-averaged network utility by jointly solving user association and power allocation under dynamic interference and self-interference conditions.
• Improve spectral efficiency and user fairness, particularly for cell-edge users, in high-density and high-traffic scenarios.

Proposed method

• Formulates the problem as a network utility maximization (NUM) problem under queue stability constraints using the Lyapunov optimization framework.
• Decomposes the NUM problem into two subproblems: auxiliary variable selection and rate maximization, which are solved in tandem.
• Models the user association and mode selection (UAMS) problem as a many-to-one matching game with pairwise stability guarantees, solved via a novel matching algorithm.
• Solves the uplink/downlink power optimization (UDPO) problem through a sequence of convex approximations using the concave-convex procedure (CCCP).
• Integrates successive interference cancellation (SIC) for NOMA decoding, with power allocation optimized per user based on channel quality and interference levels.
• Employs dynamic mode selection (FD vs. HD, NOMA vs. OMA) based on self-interference cancellation capability, network density, and traffic load.

Experimental results

Research questions

• RQ1 How can joint user association, transmission mode selection (HD/FD, OMA/NOMA), and power allocation be optimized in IBFD-NOMA networks to maximize spectral efficiency?
• RQ2 What is the impact of self-interference cancellation capability on the performance gains of IBFD-NOMA systems?
• RQ3 How does network density affect the optimal selection of transmission modes (e.g., DL NOMA vs. FD-OMA) and user scheduling?
• RQ4 To what extent can the proposed scheme improve cell-edge user throughput compared to conventional half-duplex and orthogonal multiple access baselines?
• RQ5 How does dynamic mode selection under queue stability constraints affect overall system throughput and fairness in time-varying traffic conditions?

Key findings

• The proposed scheme achieves up to 63% higher uplink packet throughput and 73% higher downlink packet throughput compared to baseline schemes.
• Cell-edge user throughput improves by 21% in uplink and 17% in downlink, demonstrating enhanced fairness and reliability.
• At low network density, coordination gain is minimal, but it increases significantly with density, reaching up to 55% gain in throughput with 14 small base stations.
• As network density increases, the system shifts from DL NOMA to FD or HD-OMA due to rising inter-cell interference, reducing the use of DL NOMA.
• The scheme maintains robust performance even at low self-interference cancellation levels (e.g., 30 dB), outperforming FD-OMA baselines due to adaptive mode switching to UL NOMA.
• At high self-interference cancellation (e.g., 110 dB), FD mode is selected more frequently, while UL NOMA usage decreases, confirming the scheme’s adaptability to hardware capabilities.