PaperlayerView exact revision on arXiv

Source-linked AI summary

Full-Duplex Wireless for 6G: Progress Brings New Opportunities and Challenges

Besma Smida, Ashutosh Sabharwal, Gabor Fodor, George C. Alexandropoulos, Himal A. Suraweera, Chan-Byoung Chae

arXiv:2304.08789v3cs.ITeess.SP

TL;DR

In-band full-duplex wireless promises higher spectral efficiency and lower latency but introduces self-interference and cross-link interference challenges. This overview surveys self-interference cancellation, antenna-based cancellation, and emerging applications including sensing, integrated access and backhaul, and reconfigurable intelligent surfaces, reporting substantial suppression results while identifying open deployment questions.

  • Problem

    In-band full-duplex wireless must manage self-interference, cross-link interference, and backscattering while supporting emerging sensing and networking applications.

  • Method

    The paper surveys self-interference cancellation techniques and examines full-duplex applications involving sensing, integrated access and backhaul, and reconfigurable intelligent surfaces.

  • Results

    Full-duplex systems achieve substantial self-interference suppression across antenna-cancellation, integrated backhaul, and RIS-assisted designs.

  • Takeaways & Limitations

    Full-duplex provides opportunities for wireless sensing, integrated sensing and communications, integrated access and backhaul, and RIS-enabled networks, alongside open research challenges.

  • Takeaways & Limitations

    Joint full-duplex transceiver and hybrid active/passive RIS design remains absent, while accurate channel acquisition and low-overhead configuration remain challenging.

Abstract

from arXiv · show

The use of in-band full-duplex (FD) enables nodes to simultaneously transmit and receive on the same frequency band, which challenges the traditional assumption in wireless network design. The full-duplex capability enhances spectral efficiency and decreases latency, which are two key drivers pushing the performance expectations of next-generation mobile networks. In less than ten years, in-band FD has advanced from being demonstrated in research labs to being implemented in standards, presenting new opportunities to utilize its foundational concepts. Some of the most significant opportunities include using FD to enable wireless networks to sense the physical environment, integrate sensing and communication applications, develop integrated access and backhaul solutions, and work with smart signal propagation environments powered by reconfigurable intelligent surfaces. However, these new opportunities also come with new challenges for large-scale commercial deployment of FD technology, such as managing self-interference, combating cross-link interference in multi-cell networks, and coexistence of dynamic time division duplex, subband FD and FD networks.

I. INTRODUCTION

In-band full-duplex has progressed from a research concept toward standardized wireless technology, while self-interference cancellation remains central to its practical operation. The paper surveys this progress and examines emerging applications and research challenges for future wireless networks.

  • Motivation: 7% annual growth in wireless broadband subscriptions and 38% mobile-network traffic growth motivate higher spectral efficiency, lower latency, and reduced infrastructure and energy costs per bit.The cited traffic reached 108 Exabytes per month, while subscriptions surpassed 7.5 billion.
  • Motivation: In-band full-duplex enables simultaneous transmission and reception in the same frequency band, unlike the traditional half-duplex assumption.Traditional networks commonly use frequency-division or time-division duplexing.
  • Progress: Within roughly one decade, in-band full-duplex moved from laboratory research into telecommunications standards and products.The transition included adoption in DOCSIS 4.0 and the appearance of FD wireless products by 2020.
  • Scope and applications: The paper surveys FD applications including wireless sensing, integrated sensing and communications, integrated access and backhaul, latency reduction, RISs, and non-terrestrial networks.It also identifies open questions and challenges for future research across these applications.
  • Technical foundation: Self-interference cancellation is the core radio technology, combining propagation, analog, and digital cancellation to manage strong self-interference.The cancellation process models and estimates the self-interference channel, reconstructs the interference, and subtracts it from the received signal; ADC dynamic range constrains the required analog cancellation.

A. Passive Analog SI Cancellation

Passive analog self-interference cancellation suppresses leakage through propagation-stage isolation, antenna-based phase cancellation, and circulators, but environmental reflections remain a major limitation.

  • Passive analog cancellation is commonly used in the propagation stage and does not require an active RF component.
  • Antenna cancellation creates a π-phase difference by combining signals from separate transmit paths with controlled path-length differences.A λ/2 distance difference between transmit antennas produces the π phase shift under line-of-sight propagation.
  • A π-phase shifter can replace spatially separated transmit antennas, and cancellation can also use two receive antennas with one transmit antenna.
  • A circulator separates transmit, antenna, and receive ports but provides isolation only over narrow bandwidths and produces reflected signal components.These components include antenna-reflected power and reflections from nearby objects.
  • Passive analog SIC cannot eliminate self-interference caused by environmental reflections, motivating combinations with shielding, directional isolation, cross-polarization, or higher-order cancellation.The cited comparison considered an anechoic chamber and a highly reflective metallic-walled room.

B. Active Analog SI Cancellation

Active analog cancellation generates a tunable signal matching residual leakage after passive suppression, using adaptive multi-tap circuits or reconfigurable impedance mismatches.

  • Active analog SIC mimics an auxiliary transmit RF chain to generate a signal matching leakage remaining after passive cancellation.A general canceller uses M delay lines, each with a tunable attenuator and phase shifter.
  • The M cancellation-parameter triplets are configured by optimizing the difference between the leaking signal and the tunable delay-line outputs.The optimization uses baseband transmit and received signals while the SI is initially transmitted at weak power to avoid ADC saturation.
  • Adaptive delay-line circuits can regenerate self-interference with arbitrary delays within a particular range by selecting appropriate fixed delays.
  • A reconfigurable impedance mismatched terminal uses circulator-reflected signals and two varactor diodes to adjust cancellation frequency band and bandwidth.The adjustment improves resilience to fabrication errors and antenna input-impedance variations.
  • More than 60 dB cancellation was achieved over 65 MHz bandwidth at 2.4 GHz using the reconfigurable IMT approach.

C. Digital SI Cancellation

Digital SI cancellation regenerates and subtracts self-interference using known transmitted data, while DNN and MIMO methods address nonlinear distortion, spatial suppression, and implementation complexity.

  • Filtering: Digital-domain SIC regenerates self-interference with filters fitted using known transmitted data.Phase noise, power-amplifier nonlinearities, and I/Q imbalance limit perfect cancellation.
  • Linearization via Deep Neural Networks (DNNs): DNN-based cancellation avoids requiring a prior nonlinear-distortion model and can reduce computational complexity by up to 36%.Current approaches often combine time-varying channel estimation with nonlinear-distortion estimation and require online training.
  • MIMO Architectures and Spatial SI Cancellation: FD MIMO uses spatial degrees of freedom to suppress SI while supporting downlink and uplink signals.Techniques include null-space projection, antenna selection, joint beam selection, lenses, and zero-forcing beamforming.
  • MIMO Architectures and Spatial SI Cancellation: The unified FD MIMO architecture connects RF chains to analog and digital beamformers, with attenuation lines and selection networks configuring analog SIC.Its architecture supports N TX antennas, M RX antennas, and configurable RF-chain counts.
  • MIMO Architectures and Spatial SI Cancellation: Fully connected analog SIC requires K = NM attenuation lines, whereas an alternative architecture uses K = N_T M_R when only selected RF chains are connected.The latter has lower complexity when N > N_T or M > M_R, but improved performance was not algorithmically justified in the cited work.
  • MIMO Architectures and Spatial SI Cancellation: The Fig. 7 architecture uses K ≤ N_T M_R active analog SIC taps, avoiding complexity that scales with the total antenna count and supporting massive MIMO.It is jointly designed with analog and digital TX/RF beamformers to address nonlinear SI.

III. CURRENT APPLICATIONS AND TRENDS

The paper surveys FD applications and trends centered on sensing, where mono-static, bi-static, and multi-static architectures trade synchronization, FD requirements, accuracy, coverage, and coordination.

  • Applications and trends: FD is presented as a technology for sensing, ISAC, IAB, latency reduction, 5G-Advanced, RIS, and NTN applications.The section highlights opportunities from STAR operation alongside application-specific open questions and challenges.
  • Wireless Networks That Sense: Higher-frequency wireless bands used for communications also support sensing applications, enabling shared hardware and infrastructure for both functions.Examples include mmWave bands and above, automotive radar, airport security, and manufacturing quality control.
  • Wireless Networks That Sense: Sensing architectures are evaluated using sensing accuracy and sensing coverage, corresponding to communication error probability and cell coverage.The three architectures differ in TX-RX placement, synchronization, FD requirements, and coordination demands.
  • Wireless Networks That Sense: Mono-static sensing co-locates TX and RX, enabling synchronization and the highest sensing accuracy, but requires in-band FD for wireless-network-scale targets.Conventional SIC can remove useful sensing information in this architecture.
  • Wireless Networks That Sense: Bi-static sensing avoids FD hardware requirements but lacks TX-RX synchronization and generally has lower sensing accuracy than mono-static sensing.Its algorithms must account for the synchronization challenge between physically separated communication ends.
  • Wireless Networks That Sense: Multi-static sensing increases coverage by using arbitrary numbers of TXs and RXs, but requires transmission coordination and joint processing of received signals.More transmitters and receivers can illuminate more of the scene and capture larger portions of reflected energy.

B. Integrated Sensing and Communication (ISAC)

ISAC jointly optimizes sensing and communication on shared hardware, with FD supporting simultaneous operation but introducing waveform, beamforming, interference, and resource-sharing challenges.

  • Integrated Sensing and Communication: ISAC jointly optimizes sensing and communication using shared hardware and a unified signal-processing framework.FD is considered enabling because its STAR capability supports both functionalities.
  • Integrated Sensing and Communication: FD ISAC research has optimized communication and radar waveforms and adapted radar processing to LTE or 5G NR time-frequency resource grids.Interpolation can address missing samples caused by null subcarriers within the transmit passband.
  • Integrated Sensing and Communication: FD systems can transmit downlink data while estimating uplink channel state information for downlink precoding in multiuser MIMO.Downlink transmission can contaminate reception of uplink training symbols, requiring training-resource optimization.
  • Integrated Sensing and Communication: FD ISAC beamformers can support multiple communication and sensing beams while suppressing SI and reducing reflections from communication beams.Beamforming nulls can be placed in frequency and angular domains, alongside OFDM waveform design.
  • Integrated Sensing and Communication: ISAC design must jointly consider sensing architecture, signal design, and resource sharing between communication and sensing operations.These choices determine how the two functions coexist in multi-function networks.

1) To Mono or Not:

Mono-static ISAC preserves sensing echoes while managing self-interference, requiring a careful shift from communication-oriented cancellation toward selective SI suppression.

  • To Mono or Not: Continuous-time mono-static sensing requires sophisticated FD operation with adequate SIC, while pulsed radar can avoid FD during silent reception periods.Known transmit signals can also support interference handling for continuous-wave operation.
  • From SIC to SI Management: Communication-oriented SIC can remove the backscattered signals needed for mono-static sensing, so ISAC must determine which reflections to suppress or admit.The BS transmits to multiple UEs while receiving echoes from surrounding objects.
  • From SIC to SI Management: Mono-static ISAC requires more than 100 dB of total SI suppression using multiple complementary methods.Useful target echoes must remain detectable while direct SI and reflections from very close surfaces are suppressed.
  • From SIC to SI Management: Radar-oriented SI management differs from communication-only FD cancellation because target echoes must be preserved rather than uniformly eliminated.This distinction requires sufficient TX/RX isolation alongside selective suppression of direct SI.

C. Integrated Access and Backhaul (IAB)

Integrated Access and Backhaul (IAB) uses wireless links to combine access and backhaul, while full-duplex IAB enables them simultaneously. Simulations and a 28 GHz prototype indicate substantial self-interference suppression, but high-power deployments remain challenging.

  • IAB integrates wireless access and backhaul, using fiber-connected IAB-donors and wirelessly forwarding IAB-nodes.
  • FD-IAB enables simultaneous backhaul and access transmission, potentially achieving more than double HD-IAB spectral efficiency while reducing end-to-end and feedback delays.
  • More than 120 dB of aggregate self-interference reduction may be necessary when an IAB-DU transmits at 46 dBm in a macrocell.
  • Link-level simulations found residual self-interference near the receiver noise floor; active analog cancellation was unnecessary at 1 and 2 meters but used at 0.1 meters.The simulations modeled a 28 GHz FD-IAB system with realistic beams, channel-aware analog cancellation, and nonlinear digital cancellation.
  • 28 GHz propagation-domain suppression averaged 100.125, 97.26, and 82.18 dB for antenna separations of 2, 1, and 0.1 meters, respectively.Directional antennas and path loss influenced the measured suppression.

D. Full-Duplex in 5G-Advanced Systems

FD schemes address traffic asymmetry and latency constraints in evolving 5G systems by supporting simultaneous uplink and downlink operation. Subband FD offers this operation across adjacent frequency resources but introduces interference challenges.

  • Conventional TDD schedules often allocate more slots to downlink than uplink, which can compromise uplink latency.
  • Dynamic TDD adapts downlink/uplink patterns to cell traffic loads but suffers from cross-link interference that can severely degrade uplink performance.
  • FD schemes can simultaneously serve uplink and downlink traffic regardless of traffic pattern and without duplexing delay on uplink traffic.
  • Subband FD assigns each TDD subband its own downlink/uplink ratio, allowing uplink and downlink on the same carrier but different frequency resources.
  • Subband FD uses adjacent TDD schedules, creating self-interference at the base station and cross-link interference between user equipment devices.
  • Subband FD faces crosstalk and adjacent-channel leakage, which can saturate or desensitize uplink receiver chains.

1) Beam Nulling:

Beam nulling uses multi-antenna spatial processing to preserve intended-link SINR while reducing interference toward non-intended receivers. In FD systems, additional nulls can isolate co-located transmit and receive antennas, but multicell interference remains difficult.

  • Beam Nulling:: Beamforming targets high SINR at intended users while creating nulls toward non-intended directions.
  • Beam Nulling:: Massive-MIMO radios supporting subband or true FD can create transmit-side nulls toward co-located receive antennas for additional isolation.
  • Beam Nulling:: Subband FD must mitigate crosstalk and adjacent-channel leakage, whereas true FD must minimize self-interference.
  • Beam Nulling:: Digital cancellation requires sufficient ADC/DAC dynamic range and resolution to capture and cancel transmit leakage alongside desired signals.
  • Beam Nulling:: RF cancellation provides a higher-dynamic-range path than digital cancellation, making it suitable for minimizing adjacent-channel leakage.
  • Beam Nulling:: RF cancellation becomes hardware-complex as the number of transmit and receive antennas grows.
  • Beam Nulling:: FD multicell networks must mitigate self-interference, cross-link interference, and conventional intercell interference, especially with high-power macrocell base stations.

E. Low-Latency Networking

FD is considered for low-latency and highly reliable applications, including mission-critical services, mixed reality, and autonomous driving. Its benefits are accompanied by interference, processing, and multi-antenna implementation challenges.

  • Low-latency and reliable communications support applications such as factory automation, telemedicine, autonomous control, and online gaming.
  • FD transmission schemes reduce end-to-end delay while maintaining high spectrum utilization and simplifying MAC design.
  • FD-based IAB can support dense wireless backhaul deployments for coverage and capacity expansion with limited capital expenditure.
  • FD relays can experience interference that limits achievable-rate gains and require higher processing power because they process twice as many packets per unit time.
  • Mixed-reality applications may benefit from simultaneous uplink and downlink for offloading and downloading, significantly reducing associated latency.
  • Vehicle-to-vehicle communication may use FD radios for simultaneous sensing and transmission at low latency without relying on cellular coverage.

F. Interplay with Reconfigurable Intelligent Surfaces (RISs)

RISs provide programmable signal control that can extend coverage, enhance desired signals, and manage interference in FD wireless systems. Their combination with FD supports two-way communication, relaying, simultaneous UL/DL operation, and new SIC designs, but joint optimization and multi-RIS coordination remain immature.

  • RIS capabilities: RISs are programmable metasurfaces that manipulate electromagnetic signals through functions such as scattering, reflection, and absorption.Their low-power unit elements are coordinated in real time by a dedicated controller.
  • Interference and efficiency: RISs can enhance desired signals or control interference, addressing the inter- and intra-cell interference escalated by overlapping FD channels.Switching RISs on or off can also combat blockage-induced effects with high energy efficiency.
  • FD–RIS use cases: RIS deployments can support FD two-way communications, relaying, and simultaneous UL/DL service while extending coverage and optimizing received SNR.Placement, orientation, propagation hops, and reflection coefficients can be designed for particular deployment objectives.
  • RIS-assisted SIC: A 256-element RIS prototype canceled 59 dB of leaked signal and achieved 103 dB total SI suppression through analog phase-reversed cancellation.The approach used a greedy heuristic to converge toward a locally optimal RIS phase profile.
  • STAR-RIS: STAR-RISs remove the conventional same-side TX/RX restriction by transmitting and reflecting incident waves toward users on both sides.Their energy-splitting, mode-switching, and time-switching protocols support diverse FD objectives; all protocols in the cited study outperformed HD and conventional RIS operation.
  • Open challenges: Joint FD–RIS research remains immature because accurate channel acquisition, low-complexity configuration, transceiver/RIS co-design, and multi-RIS coordination are unresolved.These challenges are especially relevant in multi-cell networks combining HD and FD operation.

G. Empowering Non-Terrestrial Networks (NTNs)

FD can support emerging NTN applications by improving spectral efficiency and reducing latency, while aerial-network characteristics may ease SI cancellation. Mobility, time-varying interference, shadowing, and handover create major deployment challenges.

  • 6G NTN vision: 6G NTNs aim to integrate ground and air segments through drones, HAPs, and satellite constellations to provide worldwide connectivity.The envisioned platforms include UAVs, HAPs, and low-Earth-orbit satellites operating at different altitudes.
  • SI cancellation: Shorter aerial transmission distances can reduce power imbalance between transmitted and received signals, easing SI cancellation requirements for FD operation.Powerful onboard processors, high-gain antennas, RF hardware, and low-complexity processing may further support feasible SI cancellation.
  • NTN applications: FD can benefit remote IoT services by minimizing end-to-end delivery delays and providing the higher spectral efficiencies expected by evolving NTN use cases.The cited applications include polar regions, oceans, and space.
  • Existing studies: Recent studies have examined FD communication and relaying in UAV and satellite networks, including a target of 130 dB SI cancellation for a ground–LEO link.Other reported work demonstrated more efficient satellite-spectrum use through onboard FD relaying.
  • Deployment challenges: Aerial-terminal movement complicates FD deployment through handover demands and time-varying air-to-air or air-to-ground interference.Dynamic, difficult-to-model shadowing can further complicate interference effects on performance.
  • Future outlook: Learning-based methods may support adaptive waveforms, energy-efficient operation, and SI cancellation under nonlinearities and challenging propagation conditions.The outlook places these methods in the context of 5G-Advanced Rel-18 and beyond.
Loading 2304.08789v3…