PaperlayerView exact revision on arXiv

Source-linked AI summary

Power-Domain Non-Orthogonal Multiple Access (NOMA) in 5G Systems: Potentials and Challenges

S. M. Riazul Islam, Nurilla Avazov, Octavia A. Dobre, Kyung Sup Kwak

arXiv:1609.06261v1cs.ITcs.NI

TL;DR

NOMA addresses the need for more efficient 5G radio access as networks pursue higher data rates and lower latency. This paper surveys power-domain NOMA and its research solutions, performance, integrations, and implementation issues. The surveyed evidence covers capacity, power allocation, fairness, pairing, and multiple wireless-system integrations, alongside practical limitations and research needs.

  • Problem

    5G requires enhanced radio access to support high network and user data rates, low latency, and improved spectral efficiency.

  • Method

    The paper comprehensively surveys NOMA capacity analysis, power allocation, fairness, user pairing, integrated wireless techniques, performance, challenges, and implementation issues.

  • Results

    The survey reports NOMA rate-region advantages over OMA in asymmetric channels and examines performance across cooperative, MIMO, beamforming, network, and other NOMA solutions.

  • Takeaways & Limitations

    NOMA is presented as a promising 5G access approach whose practical development requires attention to interference, user pairing, channel rank, and implementation conditions.

Abstract

from arXiv · show

Non-orthogonal multiple access (NOMA) is one of the promising radio access techniques for performance enhancement in next-generation cellular communications. Compared to orthogonal frequency division multiple access (OFDMA), which is a well-known high-capacity orthogonal multiple access (OMA) technique, NOMA offers a set of desirable benefits, including greater spectrum efficiency. There are different types of NOMA techniques, including power-domain and code-domain. This paper primarily focuses on power-domain NOMA that utilizes superposition coding (SC) at the transmitter and successive interference cancellation (SIC) at the receiver. Various researchers have demonstrated that NOMA can be used effectively to meet both network-level and user-experienced data rate requirements of fifth-generation (5G) technologies. From that perspective, this paper comprehensively surveys the recent progress of NOMA in 5G systems, reviewing the state-of-the-art capacity analysis, power allocation strategies, user fairness, and user-pairing schemes in NOMA. In addition, this paper discusses how NOMA performs when it is integrated with various proven wireless communications techniques, such as cooperative communications, multiple input multiple output (MIMO), beamforming, space time coding, and network coding, among others. Furthermore, this paper discusses several important issues on NOMA implementation and provides some avenues for future research.

I. INTRODUCTION

The paper motivates power-domain NOMA as a 5G access technique for demanding data-rate and latency targets, then surveys its concepts, research solutions, performance, implementation challenges, and related wireless integrations.

  • I. INTRODUCTION: 5G targets network-level data rates of 10-20 Gbps, user-experienced rates of 1 Gbps, and 1 millisecond end-to-end latency.
  • I. INTRODUCTION: NOMA improves spectral efficiency through superposition coding at the transmitter and successive interference cancellation at the receiver.
  • I. INTRODUCTION: The survey reviews NOMA performance analysis, fairness, energy efficiency, user pairing, solutions, implementation challenges, and integration with existing wireless technologies.
  • II. BASIC CONCEPTS OF NOMA: NOMA techniques are broadly classified into power-domain and code-domain approaches, with code-domain methods using user-specific spreading sequences.
  • A. Superposition Coding (SC): Superposition coding simultaneously transmits multiple users’ information, while the transmitter assigns different power levels to their encoded signals.

B. Successive Interference Cancellation (SIC)

SIC decodes superposed user signals successively by exploiting signal-strength differences and subtracting decoded signals before subsequent decoding. In downlink NOMA, this mechanism supports simultaneous resource use and rate advantages over OMA, while imperfect cancellation remains a concern.

  • B. Successive Interference Cancellation (SIC): Receivers order users according to signal strength so that one decoded signal can be removed before recovering another user’s message.
  • B. Successive Interference Cancellation (SIC): SIC successively decodes user signals, subtracting each decoded signal from the combined signal before decoding the next one.
  • C. A Typical NOMA Scheme: In the two-user downlink example, the stronger-channel user decodes and subtracts the other user’s signal before decoding its own.
  • C. A Typical NOMA Scheme: NOMA uses the full 1 Hz bandwidth simultaneously for two users, whereas OMA divides that bandwidth between them.
  • B. Successive Interference Cancellation (SIC): Imperfect SIC can propagate decoding errors into subsequent NOMA-user decoding stages.
  • C. A Typical NOMA Scheme: For asymmetric channels, NOMA rate pairs can exceed the corresponding OMA rates, and its achievable-rate boundary lies outside the OMA capacity region.

III. POTENTIAL NOMA SOLUTIONS

This section surveys potential NOMA solutions for problems associated with integrating NOMA into 5G, emphasizing core ideas rather than detailed derivations.

  • III. POTENTIAL NOMA SOLUTIONS: The section categorizes present and emerging NOMA research as potential solutions to integration problems and issues in 5G systems.
  • III. POTENTIAL NOMA SOLUTIONS: The survey avoids detailed explanations and mathematical derivations, directing readers to original articles for greater depth.

A. Impact of Path Loss

For randomly deployed users, NOMA path-loss performance is evaluated through outage probability for QoS-constrained rates and ergodic sum rate for opportunistically allocated rates.

  • A. Impact of Path Loss: NOMA path-loss performance is assessed under fixed target rates and opportunistic rate allocation scenarios.Outage probability measures QoS satisfaction in the first scenario, while ergodic sum rate evaluates the second.
  • A. Impact of Path Loss: Properly chosen data rates and power allocation can give NOMA better outage performance than other OMA techniques.
  • A. Impact of Path Loss: NOMA can achieve a superior ergodic sum rate when users’ data rates and assigned power are selected appropriately.
  • A. Impact of Path Loss: At high SNR, the outage probability of a user in a disk-shaped cell can be expressed analytically.The supplied passage introduces the expression but does not preserve its complete formula.
  • A. Impact of Path Loss: Calculating outage probabilities involves a complexity–accuracy trade-off that matters for precise resource allocation.

B. Cooperative NOMA (C-NOMA)

Cooperative NOMA combines relaying with NOMA to study outage, spectral efficiency, and fairness, with performance depending on relay placement and channel conditions.

  • B. Cooperative NOMA (C-NOMA): Cooperative communications with NOMA can improve system efficiency by using relay nodes to forward source information.
  • B. Cooperative NOMA (C-NOMA): Relay location determines whether NOMA or conventional OMA has better outage performance in multiple-antenna amplify-and-forward relay networks.NOMA performs better when the relay is close to the base station, whereas OMA performs better when it is close to users.
  • B. Cooperative NOMA (C-NOMA): NOMA provides better spectral efficiency and user fairness in the considered multiple-antenna relay network.
  • B. Cooperative NOMA (C-NOMA): NOMA-based cooperative relaying achieves more spectral efficiency than conventional cooperative relaying at high SNR under favorable source-to-relay channel conditions.The condition is that average source-to-relay channel power exceeds source-to-destination and relay-to-destination channel powers.
  • B. Cooperative NOMA (C-NOMA): Fairness can be addressed through power allocation, outage-probability minimization, or scheduling of concurrent transmissions.

D. NOMA with MIMO and Beamforming

NOMA integrated with MIMO and beamforming can improve capacity, outage performance, spectral efficiency, and fairness, but its gains depend on user correlation, interference, CSI, and precoder design.

  • D. NOMA with MIMO and Beamforming: NOMA beamforming lets two users share one beamforming vector while using clustering and power allocation to manage inter- and intra-beam interference.Clustering uses user correlation, while power allocation uses channel-gain differences.
  • D. NOMA with MIMO and Beamforming: NOMA beamforming improves sum capacity relative to conventional multi-user beamforming and guarantees capacity for weak users.
  • D. NOMA with MIMO and Beamforming: The beamforming clustering approach degrades considerably when too few user sets have highly correlated channels because multi-user interference increases.
  • D. NOMA with MIMO and Beamforming: MIMO NOMA can outperform conventional MIMO OMA in throughput and outage performance, including for users experiencing strong co-channel interference.
  • D. NOMA with MIMO and Beamforming: MIMO NOMA with one shared precoder per cluster can limit weak-user rates because both users’ SINRs affect decoding.Different precoders for the two users can further enhance the sum rate.
  • D. NOMA with MIMO and Beamforming: CoMP-NOMA using Alamouti coding avoids instantaneous CSI exchange, reducing backhaul overhead for high-mobility cell-edge users.

F. Network NOMA

Network NOMA extends single-cell NOMA to multi-cell settings, where inter-cell and mutual interference must be mitigated. Joint precoding is a possible solution, but its design requires broad user and channel information and may not support multiple spatially separated users.

  • Network interference: In a two-cell NOMA network, cell-edge users cannot perform SIC before decoding and therefore experience interference from neighboring-cell transmissions.The described setup pairs two users in each cell, with inter-cell interference in downlink and mutual interference in uplink transmissions.
  • Network NOMA: Single-cell NOMA solutions are insufficient for multi-cell scenarios, motivating an extension to network NOMA.The paper identifies interference mitigation as necessary when NOMA is deployed across cells.
  • Interference mitigation: Joint precoding across neighboring cells can mitigate interference, but optimal design is difficult because each base station must know all users’ data and CSI.Correlation-based precoding also requires dynamic user selection for each NOMA pair.
  • Precoding: The zero-forcing precoder is defined as the normalized pseudo-inverse of the channel-related matrix.The passage introduces the channel vectors and channel responses used in the precoder construction.

G. NOMA User Pairing

NOMA user pairing and grouping manage interference-limited operation while seeking fairness and performance gains. The surveyed approaches include channel-distinctive grouping, proportional-fairness scheduling, cooperative extensions, coding methods, and coexistence with OMA.

  • User grouping: Because NOMA is interference-limited, users are divided into groups with NOMA within groups and OMA between groups.Performance depends on which users are grouped together.
  • User pairing: Grouping users with more distinctive channel conditions can increase NOMA’s performance gain over conventional OMA under fixed power allocation.A proportional-fairness pairing method optimizes power allocation for candidate sets before scheduling the best pair or single user.
  • Fair scheduling: The UPPA approach reduces computation by avoiding unnecessary comparisons of candidate user pairs through pairing prerequisites and scheduling metrics.Prerequisite checking can omit candidate pairs whose channel or SINR conditions are unsuitable.
  • MIMO grouping: MIMO NOMA’s performance gain over MIMO OMA increases as the number of users in each group increases.This result connects grouping decisions directly to the relative performance of MIMO NOMA.
  • OMA coexistence: OMA and NOMA are expected to coexist because OMA may be preferable in small cells with few users and an unimportant near–far effect.The paper does not present NOMA as a universal replacement for OMA.

IV. PERFORMANCE EVALUATIONS

The paper evaluates NOMA using outage probability, achievable rate, and broader spectral- and energy-efficiency perspectives across cellular, cooperative, and MIMO configurations. Reported comparisons generally favor NOMA under the stated high-SNR or configuration conditions.

  • Evaluation metrics: Outage probability and achievable rate are the primary metrics used to evaluate varied NOMA setups, alongside spectral- and energy-efficiency insights.The evaluation covers several NOMA configurations rather than a single deployment model.
  • Outage probability: At sufficiently high SNR, NOMA outperforms the comparable OMA scheme in outage performance and exhibits diversity orders determined by users’ channel conditions.With a 1:4 strong-to-weak-user power allocation ratio, low-SNR outage performance is poorer for the strong user before power-domain multiplexing dominates.
  • Cooperative NOMA: Cooperative NOMA achieves maximum diversity gain for all users and transcends non-cooperative NOMA in outage performance.The comparison uses the same number of users and power allocation ratio as the preceding outage evaluation.
  • MIMO NOMA: MIMO NOMA outperforms MIMO OMA particularly at high SNR in the evaluated two-cluster, three-antenna-per-user configuration.Each cluster contains two users, with a 1:4 power allocation ratio and target rates of 3 BPCU and 1.3 BPCU.

B. Achievable Rate

Achievable-rate evaluations compare NOMA variants with OMA-based or conventional alternatives across cellular, relay, beamforming, coding, pairing, and optical settings. The reported results consistently identify higher rates, capacity, or performance gains for the evaluated NOMA approaches under their stated conditions.

  • Cellular sum rate: NOMA achieves a higher sum rate than OMA across the entire evaluated SNR range for randomly deployed two-user cellular systems.User rates are assigned opportunistically according to channel conditions, with the higher-gain user receiving the higher data rate.
  • Cooperative relaying: Cooperative relaying with NOMA achieves better average rate performance than usual relaying, particularly at high SNR when the source-to-relay channel is stronger.The result supports selecting relays from cell-center users in the evaluated NOMA relay setting.
  • Beamforming: NOMA beamforming improves sum capacity over conventional multi-user beamforming by using correlation-based clustering and effective power allocation.Sharing one beamforming vector between two users can increase the number of supportable users.
  • User pairing: PF-based user pairing provides better performance gain relative to OFDMA than TTPA across the evaluated numbers of users per cell.Prerequisite checking may partly contribute by removing invalid user pairs.
  • Visible-light communications: NOMA-based visible-light communication produces higher rate pairs than OFDMA-based VLC when power coefficients or bandwidth assignments are varied.The evaluated capacity regions support power-domain multiplexing as a promising approach for light communications.

C. Spectral Efficiency (SE) and Energy Efficiency (EE)

Network NOMA outperforms conventional single-cell NOMA in total spectral efficiency, while MIMO NOMA offers a better energy-efficiency–spectral-efficiency trade-off than OMA. The section also identifies interference, resource allocation, and distortion as important constraints on these gains.

  • Network NOMA outperforms conventional single-cell NOMA in total spectral efficiency through coordinated transmission to cell-edge users.Cell-center users detect and subtract cell-edge signals to mitigate mutual interference.
  • NOMA provides a better energy-efficiency–spectral-efficiency trade-off than OMA in single-user MIMO under statistical CSI.Both energy efficiency and spectral efficiency at the peak operating point are higher for NOMA, with larger differences at high SNR.
  • Strong co-channel interference makes joint NOMA operation difficult, motivating grouping with orthogonal bandwidth allocation between groups.Dynamic user pairing and grouping are necessary to pursue NOMA’s maximum benefits, although the cited UPPA scheme schedules only two users.
  • NOMA outage-rate optimization may not minimize expected source distortion over fading channels.The passage identifies outage-probability optimization for acceptable distortion as an open direction.

C. Impact of Interference

Interference is a central challenge for NOMA across cooperative, resource-management, MIMO, millimeter-wave, and antenna-selection settings. The surveyed approaches address interference and hardware constraints, but several deployment conditions remain important boundaries.

  • Bluetooth interference in cooperative NOMA can decrease coverage and throughput and cause intermittent or complete connectivity loss.The issue arises from interference with existing wireless personal area network operations.
  • NOMA resource management requires sophisticated user pairing and power-allocation algorithms to achieve high performance with minimal resources.Joint power and channel allocation must account for user power control and SIC implementation.
  • Existing MIMO NOMA studies commonly assume full-rank channel matrices, although channel rank determines the number of decorrelated channels and system performance.Under the full-rank constraint, cited analyses provide upper bounds for outage probability and capacity.
  • In millimeter-wave cellular systems, most locations experience signal outage beyond 175 m, with actual outage potentially increasing because of local obstacles.The passage identifies uniform outage for users beyond 150 m toward the cell boundary as an open NOMA objective.
  • Transmit antenna selection reduces complexity, power consumption, cost, and size by using one RF chain, at the expense of acceptable performance loss.The surveyed direction is a TAS-NOMA downlink scheme that improves sum rate through opportunistic channel-based allocation.

K. Carrier Aggregation

NOMA implementation involves practical constraints in signaling, processing, decoding, hardware, and optimization. These constraints motivate grouping, stronger coding, nonlinear detection, and more efficient allocation strategies.

  • SIC complexity increases with the number of users, motivating clusters containing small numbers of users for group-wise SC and SIC.Receivers must decode other users’ information before decoding their own.
  • SIC error propagation can have a marginal impact on NOMA performance under a cited worst-case simulation model.Stronger coding and nonlinear detection are proposed to compensate for propagation effects.
  • Imperfect SIC remains insufficiently understood mathematically, making its impact on NOMA performance an open research direction.Existing analytical work is noted for basic MIMO systems rather than prominent general NOMA analysis.
  • Large received-power disparities force ADCs toward high resolution, but arbitrarily high resolution is limited by cost, conversion time, and hardware complexity.
  • Exhaustive search over user pairs with dynamic power allocation is computationally expensive despite its role in maximizing throughput.Power allocation for one user also affects the achievable capacity of other power-domain multiplexed users.

E. Residual Timing Offset

Practical NOMA deployment must address asynchronous uplink timing, power-balance conditions, limited pairing opportunities, and standardization requirements. The survey presents NOMA as a candidate technology while identifying implementation boundaries and research needs.

  • Perfect synchronization is impractical for uplink NOMA because users are spatially distributed and channels are dynamic.Time-misaligned OFDM symbols from superposition-coded users can affect NOMA-user performance.
  • IDMA is robust in power-balanced scenarios with higher computational complexity, whereas NOMA is effective in power-imbalanced scenarios.
  • NOMA user pairing is limited because pairing a cell-edge user with a cell-center user requires about an 8 difference in propagation loss.The passage identifies improved detection and decoding strategies as necessary to increase the number of user pairs.
  • NOMA gains remain achievable in small cells, where NOMA offers higher performance gain than OMA.
  • 3GPP LTE Release 13 approved a study item on downlink multiuser superposition transmission to evaluate enhanced within-cell multiuser transmission.
  • The paper surveys present and emerging SC-based NOMA research in 5G and reports results concerning outage, capacity, weak-user rates, and cell-edge experiences.Its stated scope includes performance and implementation issues rather than detailed mathematical derivations.
Loading 1609.06261v1…