PaperlayerView exact revision on arXiv

Source-linked AI summary

Non-Orthogonal Multiple Access for 5G and Beyond

Yuanwei Liu, Zhijin Qin, Maged Elkashlan, Zhiguo Ding, Arumugam Nallanathan, Lajos Hanzo

arXiv:1808.00277v1cs.IT

TL;DR

NOMA research addresses energy-constrained wireless-device operation and the need for unified system frameworks. This article surveys power-domain NOMA across multiple antennas, cooperative communications, resource control, and coexistence with other 5G techniques, highlighting bandwidth efficiency, energy efficiency, fairness, and spatial-diversity considerations.

  • Problem

    Energy consumption is a critical factor for energy-constrained wireless devices, while a unified NOMA framework is sought.

  • Method

    The article surveys power-domain multiplexing aided NOMA, covering basic principles, multiple-antenna techniques, cooperative communications, resource control, and coexistence with key 5G techniques.

  • Results

    The survey highlights NOMA's bandwidth efficiency, energy efficiency, fairness, and the extra spatial diversity available through multiple-antenna techniques.

  • Takeaways & Limitations

    Future NOMA scheduling should account for the recommendations developed in the paper while considering the performance of the entire network.

  • Takeaways & Limitations

    Joint optimization of user scheduling and power allocation remains a non-trivial problem.

Abstract

from arXiv · show

Driven by the rapid escalation of the wireless capacity requirements imposed by advanced multimedia applications (e.g., ultra-high-definition video, virtual reality etc.), as well as the dramatically increasing demand for user access required for the Internet of Things (IoT), the fifth generation (5G) networks face challenges in terms of supporting large-scale heterogeneous data traffic. Non-orthogonal multiple access (NOMA), which has been recently proposed for the 3rd generation partnership projects long-term evolution advanced (3GPP-LTE-A), constitutes a promising technology of addressing the above-mentioned challenges in 5G networks by accommodating several users within the same orthogonal resource block. By doing so, significant bandwidth efficiency enhancement can be attained over conventional orthogonal multiple access (OMA) techniques. This motivated numerous researchers to dedicate substantial research contributions to this field. In this context, we provide a comprehensive overview of the state-of-the-art in power-domain multiplexing aided NOMA, with a focus on the theoretical NOMA principles, multiple antenna aided NOMA design, on the interplay between NOMA and cooperative transmission, on the resource control of NOMA, on the co-existence of NOMA with other emerging potential 5G techniques and on the comparison with other NOMA variants. We highlight the main advantages of power-domain multiplexing NOMA compared to other existing NOMA techniques. We summarize the challenges of existing research contributions of NOMA and provide potential solutions. Finally, we offer some design guidelines for NOMA systems and identify promising research opportunities for the future.

I. INTRODUCTION

Wireless communication evolved through successive standardization generations and increasingly flexible multiple-access and antenna technologies. The introduction frames 5G as a response to sharply rising traffic, connectivity, and application demands.

  • Multiple-access evolution: OMA schemes including FDMA, TDMA, CDMA, and OFDMA were complemented by spatial, antenna-array, and NOMA developments.The roadmap describes OMA and NOMA as converging with spatial-domain and multifunctional antenna technologies.
  • 5G technologies: 5G research combines ultra-densification, millimeter-wave communications, massive MIMO, D2D, full duplex, energy harvesting, C-RAN, virtualization, and software-defined networking.These directions are presented as candidate technologies within the emerging 5G network structure.

B. State-of-the-art of Multiple Access Techniques

Multiple-access techniques evolved from orthogonal resource partitioning toward NOMA, which multiplexes users within the same resource block. This survey addresses gaps in the analytical coverage, applications, comparisons, and research outlook of power-domain NOMA.

  • Multiple-access foundations: OMA assigns a single user to each time/frequency resource block, whereas CDMA supports multiple users through distinct spreading sequences.The section categorizes multiple-access techniques according to whether users share the same time or frequency resource.
  • NOMA variants: NOMA is classified mainly into code-domain and power-domain forms, with spatial-domain access also identified as another multiplexing domain.Representative code-domain methods include IDMA, LDS-CDMA, MUSA, PDMA, and SCMA.
  • Power-domain NOMA: Power-domain NOMA has a superior capacity region compared to OMA by serving multiple users in one resource block through superposition coding and SIC.Its design exploits power-domain multiplexing at the transmitter and successive interference cancellation at the receiver.
  • System-level rationale: NOMA can exploit users’ channel conditions to serve multiple users with different QoS requirements in the same resource block.The section also notes that power-domain NOMA can integrate with existing multiple-access paradigms.
  • Research gap: Existing literature had not comprehensively covered important NOMA models, analytical foundations, applications, historical milestones, and comparisons with other practical NOMA forms.The survey was developed to address these omissions and cover major NOMA challenges, opportunities, techniques, and application scenarios.

D. Organization

The paper is organized from NOMA fundamentals and multi-user detection through antenna, cooperative, resource-control, coexistence, implementation, and variant-specific topics. Its technical overview emphasizes multiplexing, interference cancellation, and their trade-offs.

  • Paper organization: Sections II–IX cover NOMA principles, antenna-aided transmission, cooperative transmission, resource allocation, coexistence with 5G technologies, implementation, variants, and conclusions.The organization follows the paper’s stated progression from fundamentals to applications, challenges, and standardization.
  • NOMA fundamentals: NOMA multiplexes users in the power domain within a conventional time/frequency domain and is presented as an add-on to legacy multiple-access solutions.The section introduces this concept together with multi-user detection, interference cancellation, and single-antenna downlink and uplink studies.
  • Multi-user detection: Multi-user detection uses users’ signal information to improve capacity, while interference-cancellation receivers reduce multiple-access interference at a complexity cost.The discussion surveys maximum-likelihood, turbo, SIC, PIC, and related receiver strategies.
  • Interference limitation: Multiple-access interference can limit capacity and performance as the number or power of interferers increases.This limitation is stated for CDMA-style systems and motivates multi-user detection and interference cancellation.
  • Interference-cancellation trade-off: PIC outperforms SIC when received signals are equalized, whereas SIC performs better when received powers differ.The stated reason is that SIC can detect the strongest user’s signal first under unequal received powers.

B. Key Technologies of NOMA

NOMA combines superposition coding at the transmitter with successive interference cancellation at the receiver, multiplexing users in the power domain on shared resources. Its analytical formulation compares NOMA and OMA throughput under channel, power-allocation, and resource constraints.

  • NOMA principles: NOMA uses superposition coding to combine user signals and SIC to cancel interference at receivers.Power-domain NOMA superimposes signals in the same time-and-frequency resource at different power levels.
  • Superposition coding: SC encodes weaker-channel users at lower rates before superimposing signals for users with better channel conditions.SC is theoretically capable of approaching Gaussian and general broadcast-channel capacities.
  • Successive interference cancellation: SIC detects the strongest signal first, re-encodes and subtracts it, and repeats until the weakest user decodes without interference.This iterative procedure differs from joint multiuser detection and supports sequential interference removal.
  • Successive interference cancellation: SIC can approach broadcast- and multiple-access-channel capacity boundaries and performs better than PIC when received powers differ.Its iterative operation also imposes lower receiver hardware complexity than joint decoding.
  • NOMA versus OMA: Power-domain NOMA allocates multiple users to one time-and-frequency resource, whereas OMA assigns resources orthogonally through allocation coefficients.The formulation considers user channel coefficients, transmit SNR, power-allocation coefficients, and resource allocation.
  • NOMA versus OMA: For two users with |h_m|^2 < |h_n|^2, NOMA has higher sum throughput than OMA, with larger gains as channel conditions become more different.The comparison assumes high SNR and equal time/frequency-resource allocation per user.

D. Main Advantages of NOMA

NOMA’s principal advantages are improved bandwidth and spectral efficiency, fairness, connectivity, compatibility, and flexibility. Studies across downlink, uplink, fading, and related communication settings report performance gains over OMA under various conditions.

  • High bandwidth efficiency: NOMA improves bandwidth efficiency and throughput by allowing multiple users to exploit each resource block.This contrasts with orthogonal allocation, where users do not simultaneously share the same time/frequency resource.
  • Fairness: NOMA allocates more power to weak users, supporting a tradeoff between user-throughput fairness.Power-allocation policies and cooperative NOMA are identified as fairness-maintenance techniques.
  • Ultra-high connectivity: NOMA can serve many IoT devices with fewer resource blocks than OMA, which requires the same number of blocks as devices.The non-orthogonal resource structure is presented as an alternative for ultra-high connectivity.
  • Compatibility: NOMA can function as an add-on to TDMA, FDMA, CDMA, and OFDMA by exploiting the power domain alongside existing access dimensions.The paper connects this compatibility to the maturity of superposition coding and SIC.
  • Flexibility: Compared with MUSA, PDMA, and SCMA, NOMA offers a conceptually appealing, low-complexity design based on allocating multiple users to one resource block.SCMA is described as a NOMA variation integrating coding, modulation, and subcarrier allocation.
  • Performance gain: NOMA’s capacity region can exceed OMA’s, while reported gains include higher downlink throughput, improved fairness, and better energy efficiency.Fading-channel results also report superior average sum-rate under minimum individual-rate constraints.

G. Uplink NOMA Transmission

Uplink NOMA has users transmit simultaneously to the base station, where SIC detects and subtracts signals sequentially. Its gains and design issues differ from downlink NOMA, especially in power control, fairness, ordering, and multi-antenna implementation.

  • Uplink operation: In uplink NOMA, multiple users transmit on the same resource and the base station detects their messages using SIC.The base station detects a stronger signal while treating another signal as interference, then subtracts the recovered signal.
  • Transmit power: Uplink users need not transmit at different powers because received SINR differences can arise from their channel conditions.When channel conditions differ substantially, received powers at the base station may differ regardless of transmit power.
  • SIC operations: Uplink and downlink NOMA differ in SIC placement and interference handling: uplink cancellation occurs at the base station, whereas downlink users perform SIC.These distinctions are explicitly tied to the respective transmission directions.
  • Performance gain: NOMA’s capacity region lies outside OMA’s, with downlink gains emphasized for throughput and uplink gains emphasized for fairness.Reported uplink benefits over OMA become more significant when users’ channel conditions are more different.
  • Research directions: Uplink research covers power control, outage probability, delay-limited sum-rate, energy harvesting, user pairing, and stochastic-geometry models.The cited studies include OFDMA uplink extensions, iterative detection, wirelessly powered transmission, and general MIMO-NOMA frameworks.
  • Open challenges: Open challenges include user ordering for SIC, imperfect CSI, limited feedback, receiver design, adaptive coding, and closed-form multi-cell analysis.Multiple-antenna NOMA further complicates power-based ordering because channels are matrices, making beamforming or precoding essential.

A. Cluster-Based MIMO-NOMA

Cluster-based MIMO-NOMA organizes users into clusters and combines beamforming, precoding, detection, and SIC to manage interference and heterogeneous QoS requirements. The surveyed designs improve performance, fairness, or scalability, but depend on channel-state information and face feedback and complexity challenges.

  • Cluster-based design: Cluster-based MIMO-NOMA partitions users into clusters and can decompose the system into independent SISO-NOMA arrangements.This structure supports beamforming and SIC within clusters while managing inter-cluster interference.
  • Interference management: Transmit precoding, detector design, user allocation, and dynamic power allocation can make cluster beams orthogonal to other clusters’ channels, efficiently suppressing inter-cluster interference.SIC remains available for intra-cluster interference when users within a cluster have different channel conditions.
  • QoS-aware clustering: QoS-based user grouping avoids the restrictive assumption that different users must have different channel conditions.The design compares users grouped by QoS requirements against channel-ordered ZF-NOMA and SA-NOMA benchmarks.
  • QoS-aware clustering: QR-based MIMO-NOMA exploits heterogeneous QoS requirements and outperforms ZF-NOMA, SA-NOMA, and MIMO-OMA in outage probability.The scheme uses QR decomposition to increase differences between effective channel conditions.
  • Beamformer-based design: Layered transmission produces an achievable sum rate that increases linearly with the number of antennas.Beamformer weights are calculated in an order beginning with the most demanding QoS requirement, while users share resource blocks and apply SIC.
  • Massive-MIMO-NOMA: Massive-MIMO-NOMA can reduce feedback overhead through one-bit CSI feedback, but its performance is degraded relative to perfect CSIT.Obtaining CSIT is difficult in massive-MIMO-NOMA because pilot overhead and computational complexity grow with the antenna count and Doppler frequency.

IV. INTERPLAY BETWEEN NOMA AND COOPERATIVE COMMUNICATIONS

Cooperative NOMA combines SIC with decode-and-forward relaying to improve weak-user reliability, fairness, diversity, and coverage. Extensions address relaying, multi-cell interference, CoMP, and C-RAN, while introducing extra time-slot and interference-management considerations.

  • Cooperative NOMA: Cooperative NOMA uses strong NOMA users as decode-and-forward relays for weak users.The strong user forwards the decoded weak-user message during a cooperative phase.
  • Benefits: Cooperative NOMA improves weak-user reliability and fairness, particularly when the weak user is near the cell edge.SIC lets the strong user decode the weak signal before remodulating and retransmitting it closer to the destination.
  • Benefits: Cooperative NOMA provides weak users with diversity gains equal to those of conventional cooperative networks, including when energy-harvesting relays are used.The diversity benefit helps overcome multipath fading.
  • Performance comparison: Cooperative NOMA achieves higher diversity gain than non-cooperative NOMA and OMA in outage-probability comparisons.The surveyed results identify cooperative NOMA as a reliability-enhancing technique for weak users.
  • Multi-cell and CoMP NOMA: Multi-cell NOMA and CoMP use precoding, coordinated beamforming, or joint signal processing to mitigate inter-cell interference and improve cell-edge performance.CoMP enables multiple base stations to coordinate transmissions for cell-edge users.

C. Discussions and Outlook

The paper surveys cooperative, fairness-oriented, and cognitive power-allocation strategies for NOMA, emphasizing trade-offs among throughput, fairness, complexity, and QoS. It also identifies practical limitations and open directions for scalable resource control and relay design.

  • Open cooperative-NOMA issues: Cooperative relaying improves reception reliability and coverage but requires an extra time slot for relaying.Full-duplex relays are proposed to eliminate this extra slot, although self-interference and inter-user interference must then be addressed.
  • Power allocation: NOMA power allocation must balance bandwidth efficiency and energy efficiency while controlling interference and supporting weak-user fairness.Poor-channel users generally receive more power and better-channel users less power to maintain fairness.
  • Fairness strategies: Fairness objectives include ordered power allocation, max-min rate fairness, proportional fairness, α-utility, weighted sum rate, and Jain’s fairness index.The α-utility framework reduces to proportional fairness at α = 1 and approaches max-min fairness as α →∞.
  • Fairness strategies: Jain’s fairness index evaluates fairness but cannot prevent weak users from receiving low rates.The paper therefore calls for more intelligent fair power-allocation algorithms.
  • Cognitive power allocation: Cognitive power allocation is especially useful when channel ordering and power-allocation constraints are non-convex, particularly in multi-antenna NOMA.Its QoS constraints prioritize the weak user by limiting power assigned to the strong user.
  • Cognitive power allocation: Cognitive power allocation guarantees weak-user QoS while trading off overall throughput against individual fairness and reducing power-allocation complexity.It also gives the base station flexibility to exploit opportunistic support from the strong user.

B. User Scheduling in Dynamic Cluster/Pair Based Hybrid MA Networks

User scheduling and resource allocation in hybrid NOMA networks require effective pairing, clustering, power allocation, and resource-block assignment. The surveyed solutions manage computational difficulty through relaxation, monotonic optimization, matching, heuristics, and joint optimization, but often trade optimality for tractability.

  • User pairing and clustering: User pairing and clustering are central problems in hybrid multiple-access networks because users must be assigned to pairs or clusters.Pairing can improve both individual-pair and network performance while reducing design complexity.
  • User pairing and clustering: NOMA provides higher sum-rate gain over OMA when paired users have substantially different channel conditions.The result is attributed in the surveyed discussion to SIC-aided detection.
  • User pairing and clustering: User-pair secrecy diversity is determined by the user with the weaker channel, and pairing can be further improved through power allocation.The paper identifies joint pairing and power allocation as a promising direction.
  • Resource allocation complexity: Joint power and subcarrier allocation is NP-hard, making exhaustive optimization computationally prohibitive.Low-complexity suboptimal approaches are used to balance performance and complexity.
  • Resource allocation methods: Combinatorial relaxation converts binary resource-assignment variables into continuous variables, but can create a non-trivial duality gap.The relaxed problem may be solved using a Lagrangian dual method.
  • Resource allocation methods: Matching theory addresses combinatorial allocation but depends on predefined preference lists that may change as channel conditions fluctuate.This variability requires further research when the numbers of users and resource blocks are high.
  • Resource allocation methods: Heuristic algorithms provide approximate solutions at acceptable complexity, but can produce different performance-versus-complexity trade-offs.Existing work often decouples correlated scheduling and power-allocation problems, while systematic complexity analysis remains limited.

C. Software-Defined NOMA Network Architecture

The proposed SD-NOMA architecture uses a central SDN controller to coordinate distributed NOMA components and globally manage network resources. The section also surveys integration with massive MIMO, HetNets, and mmWave, emphasizing bandwidth-efficiency gains alongside interference-management challenges.

  • Software-Defined NOMA Network Architecture: The controller can jointly manage user clustering, association, and power allocation while accounting for intra-cluster and inter-cluster interference.The resulting power-allocation problem is formulated as a global optimization problem using the network-wide state.
  • Software-Defined NOMA Network Architecture: SD-NOMA uses an SDN controller with a global view of network resources and traffic to control distributed SDR-based NOMA components.This supports centralized management of user association, power allocation, interference, and clustering across the network.
  • Software-Defined NOMA Network Architecture: SD-NOMA facilitates abstractions for power optimization, interference management, user association, and dynamic user clustering or pairing.The architecture is presented as an attractive 5G network structure with flexible resource control.
  • NOMA and Massive-MIMO HetNets: A massive-MIMO HetNet framework combines macro-cell massive MIMO with user-pairing NOMA transmissions in small cells to integrate spatial reuse and bandwidth efficiency.The framework uses stochastic geometry to model and analyze a K-tier HetNet.
  • NOMA and mmWave: NOMA-mmWave systems can outperform conventional OMA-mmWave systems while enhancing bandwidth efficiency and supporting massive connectivity.NOMA is suited to mmWave beams because highly directional beams create correlated channels and suppress inter-beam interference.

C. NOMA and Cognitive Radio Networks

NOMA is surveyed alongside cognitive radio and device-to-device communications as a way to improve bandwidth efficiency through shared-resource transmission. The section identifies interference management, joint resource control, and broader standardization or implementation issues as central challenges.

  • NOMA and Cognitive Radio Networks: Cognitive radio and NOMA share interference management as a core challenge while seeking improved bandwidth efficiency.This motivates linking the two technologies, including NOMA operation in underlay cognitive-radio networks.
  • NOMA and Cognitive Radio Networks: Existing CR-NOMA studies focus on the underlay paradigm, leaving interweave and overlay CR-NOMA as open research directions.The underlay approach permits secondary access when primary-user interference constraints are satisfied.
  • NOMA and D2D Communications: NOMA-based D2D communication allows a D2D transmitter to serve multiple D2D receivers simultaneously through a D2D group.The proposed scheme was reported to deliver higher throughput than conventional D2D communication.
  • Challenges of Integrated NOMA Networks: Applying NOMA to HetNets, cognitive-radio networks, and D2D scenarios can impose increased co-channel interference on existing networks.The section therefore calls for intelligent interference management when combining these technologies.
  • Implementation Challenges: SIC suffers from inter-user error propagation because an incorrect decoded signal leaves residual interference and can corrupt subsequent users’ decoding.Many studies assume perfect interference cancellation, although inaccurate power allocation and imperfect channel decoding make that assumption difficult to satisfy in practice.

B. Channel Estimation Error and Complexity for NOMA

The section reviews practical NOMA constraints involving channel estimation, receiver complexity, security, energy harvesting, and standardization. It summarizes proposed responses including partial CSI, iterative estimation and detection, physical-layer security, SWIPT, and LTE-compatible MUST transmission.

  • Channel Estimation Error and Complexity: Channel-estimation errors create ambiguous user ordering and inaccurate power control, which affect SIC decoding accuracy.Perfect CSI is unavailable in practice, while near-optimal conventional estimation can impose excessive overhead and computational complexity, especially in MIMO-NOMA.
  • Channel Estimation Error and Complexity: Proposed responses to channel-estimation complexity include improved estimation designs, compressed sensing, partial CSI, limited feedback, and iterative joint estimation and detection.These approaches seek a complexity-performance tradeoff or reduced feedback overhead.
  • Physical-Layer Security: Physical-layer security research for NOMA includes protected zones, careful channel ordering, artificial noise, and secrecy-sum-rate optimization.One SISO-NOMA study characterized optimal power allocation in closed-form expressions.
  • Wireless Power Transfer and Energy Harvesting: SWIPT-based NOMA was proposed alongside user-selection schemes, and its analytical results confirmed no diversity-gain degradation compared with conventional NOMA.Energy harvesting is discussed as a way to support energy-constrained wireless devices and extend battery recharge periods.
  • Standardization and Implementation: NOMA standardization included LTE-A and LTE Release 13 through MUST, which can serve two users on the same OFDM subcarrier without changing the existing structure.The section notes that implementation remains challenging because receiver decoding complexity increases as the number of users grows.

F. Discussions and Outlook

The outlook identifies practical imperfections in NOMA, including synchronization, filtering distortion, and ADC limitations, while comparing power-, code-, and multi-carrier NOMA designs. It highlights trade-offs involving performance, complexity, and user support.

  • Implementation challenges: Power-domain NOMA creates ADC trade-offs because strong signals require large voltage ranges, whereas weak signals require high-resolution conversion.Practical ADCs cannot simultaneously provide both a large voltage range and high resolution, making quantization errors unavoidable.
  • Implementation challenges: The resulting design objective is to seek a suitable performance-complexity trade-off in NOMA receiver implementation.The passage frames this trade-off as necessary because ADC cost and system complexity constrain achievable conversion capabilities.
  • Implementation challenges: NOMA faces unresolved imperfections involving filtering distortion, synchronization, and the performance effects of relative time offsets among interfering users.Asynchronous-system performance was reported to depend largely on the relative time offset, while filtering distortion effects remain unknown.
  • NOMA variants: NOMA encompasses code-domain and power-domain approaches, as well as single-carrier and multi-carrier forms, whose advantages and disadvantages differ.The discussion proceeds from single-carrier NOMA toward multi-carrier designs, including OFDM coexistence.
  • Code-domain NOMA: IDMA uses user-specific chip interleavers, while LDS-CDMA uses sparse spreading and message passing to support non-orthogonal multi-user detection.IDMA can retain diversity through iterative chip-by-chip detection, whereas sparse spreading avoids the equal-users-and-chips restriction of orthogonal designs but requires sophisticated decoding.
  • Code-domain NOMA: LPMA combines power- and code-domain multiplexing with lattice coding and SIC, but has higher encoding and decoding complexity than power-domain NOMA.LPMA is also described as potentially circumventing a specific impediment of power-domain NOMA, although the supplied passage does not complete that impediment’s description.
  • Multi-carrier NOMA: Multi-carrier NOMA designs such as LDS-OFDM combine sparse spreading with OFDM and MPA detection, but mapping choices and MPA can increase complexity relative to conventional OFDMA.LDS-OFDM spreads data before OFDM modulation; alternative mappings offer different benefits, including frequency-domain diversity.

2) SCMA:

SCMA and related NOMA variants use sparse signatures, multidimensional constellations, and joint detection to support non-orthogonal access, while differing in mapping flexibility and receiver complexity. The paper surveys these schemes alongside power-domain NOMA, integration with other 5G techniques, resource control, implementation challenges, and future research needs.

  • SCMA: SCMA supports multiple users on shared subcarriers through sparse spreading codes and joint encoding across multiple subcarriers.Each user occupies only a subset of subcarriers, and receivers require joint decoding.
  • SCMA: SCMA uses multidimensional constellations and fewer constellation points at a given throughput to obtain constellation shaping gain.The approach requires sophisticated codebook design.
  • Detection: MPA-based detection can achieve near-ML performance at a fraction of maximum-likelihood complexity, with hybrid MPA-SIC designs identified as a research need.The survey also notes turbo detection and other MPA-SIC strategies for improving decoding performance.
  • PDMA: PDMA permits a variable number of users per resource block and avoids SCMA’s complex multidimensional constellation design.Its pattern matrix can assign all subcarriers to one user and only one subcarrier to another.
  • Comparative scope: Existing NOMA solutions have distinct trade-offs: SCMA suits uplink grant-free access but needs sophisticated codebooks and joint detection, whereas MUST enhances downlink bandwidth efficiency.The paper identifies a unified framework supporting numerous users in general scenarios as an open problem.
  • Survey scope: The survey covers power-domain NOMA principles, multiple-antenna and cooperative designs, resource control, coexistence with 5G techniques, implementation, and standardization.It also provides design guidelines and highlights spatial-diversity benefits from combining NOMA with multiple-antenna and cooperative techniques.
Loading 1808.00277v1…