[Paper Review] Multiple Antenna Assisted Non-Orthogonal Multiple Access

Yuanwei Liu, Hong Xing+4 more|arXiv (Cornell University)|Jan 24, 2018

Advanced Wireless Communication Technologies13 references5 citations

TL;DR

This paper investigates multiple-antenna aided non-orthogonal multiple access (MIMO-NOMA), demonstrating that MIMO-NOMA achieves higher spectral efficiency than MIMO-OMA by exploiting spatial multiplexing and beamforming. It proposes beamformer-based and cluster-based MIMO-NOMA designs, analyzes resource allocation in single- and multi-carrier systems, and identifies practical implementation challenges with solutions, highlighting key open research directions in 5G networks.

ABSTRACT

Non-orthogonal multiple access (NOMA) is potentially capable of circumventing the limitations of the classic orthogonal multiple access schemes, hence it has recently received significant research attention both in industry and academia. This article is focused on exploiting multiple antenna techniques in NOMA networks, with an emphasis on investigating the rate region of multiple-input multiple-output (MIMO)-NOMA, whist reviewing two popular multiple antennas aided NOMA structures, as well as underlining resource management problems of both single-carrier and multi-carrier MIMO-NOMA networks. This article also points out several effective methods of tackling the practical implementation constraints of multiple antenna NOMA networks. Finally, some promising open research directions are provided in context of multiple antenna aided NOMA.

Motivation & Objective

• To analyze the spectral efficiency gains of MIMO-NOMA over MIMO-OMA from an information-theoretic perspective.
• To investigate two dominant MIMO-NOMA system architectures: beamformer-based and cluster-based designs.
• To address resource allocation challenges in both single-carrier and multi-carrier MIMO-NOMA networks.
• To identify and propose solutions for practical implementation constraints such as hardware complexity and security vulnerabilities.
• To outline promising open research directions for MIMO-NOMA in future 5G and beyond networks.

Proposed method

• Uses superposition coding (SC) at the transmitter and successive interference cancellation (SIC) at receivers to multiplex users in the power domain within the same resource block.
• Analyzes the achievable rate region of MIMO-NOMA by comparing it with dirty paper coding (DPC), showing that NOMA’s rate region is a subset of DPC’s under fixed decoding order.
• Proposes beamformer-based MIMO-NOMA using directional beamforming to create user-specific channels and enhance spectral efficiency.
• Introduces cluster-based MIMO-NOMA to group users based on channel quality, enabling efficient power and beam allocation.
• Applies antenna selection (AS) to reduce RF chain costs and simplify channel ordering, transforming MIMO-NOMA into SISO-NOMA for performance-complexity tradeoff.
• Integrates physical layer security (PLS) via artificial noise (AN) injection to protect against eavesdropping, especially in scenarios with internal eavesdroppers among NOMA users.

Experimental results

Research questions

• RQ1 What is the theoretical spectral efficiency gain of MIMO-NOMA over MIMO-OMA in terms of achievable rate region?
• RQ2 How can beamforming and clustering be effectively applied in MIMO-NOMA to enhance system performance and support heterogeneous QoS requirements?
• RQ3 What are the key resource allocation challenges in single-carrier and multi-carrier MIMO-NOMA systems, and how can they be addressed?
• RQ4 How can practical implementation issues such as hardware complexity, feedback overhead, and security threats be mitigated in MIMO-NOMA?
• RQ5 What are the most promising open research directions for MIMO-NOMA in 5G and beyond, especially with massive MIMO and millimeter wave systems?

Key findings

• MIMO-NOMA achieves a larger spectral efficiency than MIMO-OMA by exploiting spatial degrees of freedom through beamforming and spatial multiplexing.
• The rate region of MIMO-NOMA is a strict subset of that of DPC, but under certain conditions—such as quasi-degradation—MIMO-NOMA can achieve DPC-level performance.
• Antenna selection (AS) enables a favorable performance-complexity tradeoff by reducing RF chains and simplifying channel ordering, effectively transforming MIMO-NOMA into SISO-NOMA.
• Physical layer security via artificial noise (AN) injection can protect against external and internal eavesdroppers, especially when users have asymmetric channel conditions.
• Stochastic geometry-based modeling using Poisson cluster processes (PCP) offers a more realistic framework for analyzing large-scale MIMO-NOMA networks with spatial randomness.
• Optimal SIC decoding order design remains an open challenge in MIMO-NOMA due to the dependence on transmit precoding and detector design, with most existing works relying on suboptimal ordering.