Source-linked AI summary
In-Band Full-Duplex Wireless: Challenges and Opportunities
Ashutosh Sabharwal, Philip Schniter, Dongning Guo, Daniel W. Bliss, Sampath Rangarajan, Risto Wichman
TL;DR
IBFD aims to increase wireless spectral efficiency by enabling simultaneous same-band transmission and reception, but self-interference and device constraints limit practical deployment. This tutorial reviews IBFD concepts, synthesizes self-interference mitigation across propagation, circuit, and digital domains, and surveys opportunities and open challenges. The paper concludes that IBFD offers numerous opportunities accompanied by challenges spanning antenna and circuit design through wireless-network foundations.
Problem
Self-interference from a terminal’s own transmission can overwhelm desired reception, while practical IBFD systems also face hardware, environmental, and network-design challenges.
Method
The tutorial reviews IBFD concepts and synthesizes self-interference suppression techniques, system opportunities, and research challenges across physical and higher network layers.
Results
IBFD offers numerous opportunities for increasing wireless-network spectral efficiency, including potentially doubling spectral efficiency in relay and bidirectional topologies.
Takeaways & Limitations
Realizing IBFD requires interdisciplinary work spanning antenna and circuit design, physical-layer algorithms, and theoretical foundations for wireless networks.
Abstract
from arXiv · showhide
In-band full-duplex (IBFD) operation has emerged as an attractive solution for increasing the throughput of wireless communication systems and networks. With IBFD, a wireless terminal is allowed to transmit and receive simultaneously in the same frequency band. This tutorial paper reviews the main concepts of IBFD wireless. Because one the biggest practical impediments to IBFD operation is the presence of self-interference, i.e., the interference caused by an IBFD node's own transmissions to its desired receptions, this tutorial surveys a wide range of IBFD self-interference mitigation techniques. Also discussed are numerous other research challenges and opportunities in the design and analysis of IBFD wireless systems.
I. INTRODUCTION
IBFD promises higher spectral efficiency and new access-layer capabilities, but self-interference and hardware limitations create substantial implementation challenges. The section motivates layered suppression techniques and broader redesign of wireless systems.
- IBFD allows terminals to transmit and receive simultaneously in the same frequency band, targeting higher wireless spectral efficiency.
- IBFD can support access-layer functions including collision detection during transmission and instantaneous feedback from other terminals.
- Self-interference can place a severe burden on full-duplex reception: in the femto-cell example, suppression exceeding 106 dB is required to match half-duplex link SNR.The calculation assumes 21 dBm transmission, 15 dB isolation, and a −100 dBm receiver noise floor.
- Limited ADC dynamic range leaves a residual self-interference floor even after perfect digital cancellation, reaching 52 dB above the equivalent half-duplex receiver noise floor.
- Self-interference mitigation combines propagation-domain, analog-circuit-domain, and digital-domain techniques, with analog cancellation preserving propagation patterns.
- Environmental effects such as nearby reflections can substantially degrade chamber suppression, motivating channel-state-information-aware digital processing.
- Commercial IBFD deployment requires reconsidering network design, while small-form-factor devices remain difficult to support experimentally.
II. LITERATURE REVIEW
IBFD has a long history in radar, where continuous-wave systems transmit and receive simultaneously using separate or shared antennas. Radar self-interference suppression evolved from antenna separation to analog cancellation.
- IBFD has been used in radar since at least the 1940s, predating its recent communications resurgence.
- Continuous-wave radars transmit and receive simultaneously, unlike pulsed radars that turn off transmission while collecting returns.
- Radar terminals can interface antennas through separate-antenna or shared-antenna full-duplex architectures.
- 1960s feed-through nulling increased continuous-wave radar dynamic range with reported 60 dB isolation, but required a 60+ kg precision ferrite canceler.
B. Research Advances in In-band Full-duplex Wireless Communications
Wireless communications historically avoided IBFD outside special cases, but relaying, bidirectional links, and broader network research have recently expanded its scope. Important physical- and higher-layer questions remain open.
- Cellular and WiFi systems largely avoided IBFD, whereas in-band relays used it to receive, amplify, and retransmit signals for coverage extension.
- IBFD relay research progressed from physical separation techniques to information-theoretic analyses incorporating residual self-interference and hardware impairments.
- Recent experiments demonstrated bidirectional IBFD over short ranges, helping drive renewed communications research beyond relay applications.
- Research now covers bottleneck characterization, transmit-chain-aware cancellation, multi-user networks, distributed full-duplex, cognitive radios, and multiband systems.
- Higher-layer work examines full-duplex medium access, cross-layer scheduling, neighbor discovery, mutual broadcasting, and ranging or localization.
- Many physical-layer and higher-layer aspects of IBFD still require more thorough investigation.
III. OPPORTUNITIES TO LEVERAGE FULL-DUPLEX
IBFD enables simultaneous same-band transmissions and receptions across relay, bidirectional, and base-station topologies, expanding network-level communication opportunities. Realizing its ideal spectral-efficiency gains requires managing both self-interference and inter-terminal interference.
- Network topologies: Relay, bidirectional, and base-station topologies illustrate how IBFD terminals can support simultaneous flows that half-duplex terminals schedule in orthogonal time slots.The relay requires only the relay to operate in full-duplex; the base-station topology similarly requires only the base station.
- Interference challenges: IBFD networks introduce both self-interference and inter-terminal interference, and ideal doubling requires managing both.Self-interference is much stronger and is identified as the limiting factor of practical IBFD networks.
- Spectral-efficiency gains: Under a peak power constraint, IBFD achieves exactly twice the half-duplex spectral efficiency at all SNRs.Under equal average power, the gain is strictly less than 2 at finite SNR and approaches 2 as SNR increases.
- Network topologies: A bidirectional network with N > 2 IBFD terminals can support up to N(N −1) simultaneous data flows.
- Higher-layer opportunities: Removing the frame-level half-duplex constraint expands protocol design opportunities, including simultaneous control-information exchange alongside data transmission.Such multiplexing could reduce protocol overhead and improve access-layer throughput.
IV. TECHNIQUES FOR SELF-INTERFERENCE REDUCTION
The paper organizes IBFD self-interference reduction techniques into three domains and distinguishes whether they actively or passively address device-extrinsic scattering.
- Self-interference reduction techniques are partitioned into propagation-domain, analog-circuit-domain, and digital-domain approaches.
- The techniques are also distinguished by whether they actively or passively mitigate self-interference caused by device-extrinsic scattering.
A. Anatomy of an In-band Full-Duplex Terminal
An IBFD terminal simultaneously transmits and receives through coordinated digital, analog, RF, antenna, and cancellation components. Its received signal contains the desired signal plus direct-path and reflected-path self-interference.
- Transmit and receive chains: The transmit chain codes and modulates digital bitstreams, converts them through DACs, upconverts and amplifies them, then radiates them through transmit antennas.Transmit processing introduces non-idealities including DAC quantization noise, oscillator phase noise, and amplifier distortion.
- Transmit and receive chains: The receive chain passes each antenna signal through an LNA, downconverter, and ADC before digital demodulation, interference cancellation, and decoding.
- Terminal architecture: A separate-antenna IBFD terminal can use multiple dedicated transmit and receive antennas, while a shared-antenna architecture pairs each transmit and receive chain through a common antenna.Shared-antenna operation requires a duplexer that routes transmit and receive signals while isolating the receiver chain.
- Self-interference paths: The received signal comprises the desired signal, direct self-interference from the transmit chain, and self-interference reflected by nearby device-extrinsic scatterers.
- Propagation-domain suppression: Propagation-domain suppression uses antenna separation, shielding, cross-polarization, directionality, and transmit beamforming to reduce self-interference before or at the antennas.Transmit beamforming adjusts per-antenna complex weights to attempt nulls at receive antennas, but pattern changes can also suppress desired signals.
- Propagation-domain suppression: Commercial hardware achieved 74 dB suppression in an anechoic chamber but only 46 dB in highly reflective indoor offices.The same design, combined with analog and digital cancellation, enabled near-perfect outdoor IBFD over ranges up to 150 meters.
C. Analog-circuit-domain Self-interference Cancellation
Analog-circuit-domain cancellation subtracts a processed copy of the transmit signal in the analog receive chain before the ADC. Its design involves tradeoffs between handling reflected-path interference and avoiding analog non-idealities.
- Analog-circuit-domain cancellation suppresses self-interference before the ADC, either before or after the downconverter and LNA.One configuration taps the transmit signal at the antenna feed, processes it analogically, and subtracts it from the receive-antenna feed.
- Channel-unaware analog cancellation targets direct-path interference, whereas channel-aware cancellation attempts to address both direct- and reflected-path interference.
- For narrowband signals, each transmit–receive antenna pair can be modeled with a complex gain and delay, reducing single-antenna cancellation to one scalar gain and one delay.Channel-unaware systems adjust these once during design or calibration, while channel-aware systems adapt them.
- Wideband reflected-path interference is frequency-selective because of multipath, making analog-circuit-domain cancellation especially challenging.Direct-path interference remains more tractable when antenna gain and phase responses are engineered to be frequency-flat.
- Tapping near the antennas avoids some analog-domain non-idealities but requires difficult analog processing for wideband reflected paths.Digital-domain tapping enables adaptive DSP for reflected interference, but downstream analog-circuit non-idealities limit cancellation precision.
D. Digital-domain Self-interference Cancellation
Digital-domain cancellation uses signal processing after the ADC, where sophisticated processing can estimate and subtract self-interference but must account for hardware and channel imperfections.
- Digital-domain cancellation: Digital-domain cancellation applies DSP after the ADC, enabling sophisticated processing such as adaptive receive beamforming.Analog receive beamforming is possible in principle but is more difficult to implement than in the digital domain.
- System modeling: Accurate discrete-time baseband models must capture propagation, analog suppression, antennas, and non-idealities between the DAC and ADC.The model may also need to include inter-antenna channels and other terminals in the network.
- Cancellation operation: A relay can estimate the partially suppressed self-interference channel, form an estimate using its intended transmit signal, and subtract it from the received signal.Channel-estimation error, transmit distortion, and receive distortion remain after this subtraction.
- Impairment mitigation: Modeling transmitter and receiver imperfections, including HPA nonlinearities, can enable their mitigation during digital cancellation.Polynomial models and transmit/receive beamforming are among the approaches described.
- Research agenda: The paper frames widespread IBFD deployment as requiring further research challenges and opportunities beyond cancellation techniques.These challenges concern the design and analysis of future wireless networks.
A. Antenna and Circuit Design
IBFD creates design opportunities across antennas, circuits, physical-layer algorithms, and networks, but practical deployment must balance suppression, device constraints, and protocol complexity.
- Antenna and circuit design: Small-form-factor devices make antenna design a major IBFD challenge because propagation-domain and analog-circuit-domain suppression depend heavily on antenna performance.The paper highlights the need to circumvent the impact of small apertures on IBFD performance.
- Antenna and circuit design: IBFD circuitry must balance self-interference cancellation against device cost, physical area, and power consumption.Analog-circuit limitations and oscillator phase noise can constrain cancellation performance and spectral-efficiency gains.
- Physical-layer algorithm design: Accurate statistical characterization of the effective channel is necessary because it includes antennas, circuitry, propagation, non-idealities, suppression, noise, and external interference.Residual interference can often be modeled as Rician, while its coherence interval and bandwidth depend on multipath and mobility.
- Physical-layer algorithm design: IBFD resource allocation must jointly consider space, time, and frequency because suppression can alter desired-signal power and channel conditions vary with frequency.The paper identifies joint space-time-frequency allocation as an important research problem.
- Physical-layer algorithm design: Fundamental IBFD limits should be studied using capacity, outage capacity, diversity-multiplexing tradeoff, and network-throughput metrics with faithful impairment models.The value of these analyses depends on the fidelity of their propagation and circuit assumptions.
- Network foundations and protocol design: IBFD expands higher-layer protocol design by removing the frame-level half-duplex constraint, potentially enabling new throughput-oriented mechanisms.The paper discusses MAC, routing, multiuser decoding, and virtual-IBFD alternatives when cancellation is infeasible or too expensive.
- Network foundations and protocol design: The capacity advantage of IBFD in general wireless networks remains an open problem, motivating scenario-specific characterization for practical guidance.This limitation applies to wireless networks with or without IBFD.
VI. CONCLUSIONS
IBFD offers opportunities to increase wireless-network spectral efficiency, but realizing them requires interdisciplinary work across antenna, circuit, protocol, and theoretical design.
- VI. CONCLUSIONS: IBFD offers numerous opportunities for increasing wireless-network spectral efficiency while presenting challenges across all system layers.The paper emphasizes that substantial research remains before these opportunities are fully realized.
APPENDIX
The appendix analyzes how ADC quantization and dynamic range constrain SINR after digital self-interference cancellation, including the effects of ENOB, noise, and signal PAPR.
- ADC model: The ADC output model adds quantization error e to the input x, whose components are desired signal d, self-interference s, and noise n.The input is modeled as x = d + s + n and the output as y = x + e.
- ADC model: The ENOB-dependent term converts ADC resolution into quantization-error power, while PAPRx accounts for the peak level needed to prevent clipping.Each additional ENOB reduces quantization-error amplitude by approximately 6.02 dB.
- Digital cancellation: Perfect digital cancellation forms z = y − s = d + n + e, leaving desired signal, noise, and quantization error in the received signal.The resulting SINR treats quantization error as uncorrelated with noise.
- SINR analysis: Under the stated ENOB choice, quantization error power is 6.02 dB, or 1 bit, below the noise power before the SINR bound is derived.The assumption is expressed as quantization-error variance being one quarter of the noise variance.
- SINR analysis: After perfect digital cancellation, SINR is upper bounded by pre-ADC signal-to-interference ratio plus ADC dynamic range and a PAPR-dependent penalty.The bound is written in dB and applies to the self-interference-dominated case.
- SINR analysis: 5 dB PAPRx produces a −12 dB penalty, equivalent to a loss of 2 bits relative to ENOB.The appendix also states an effective ADC dynamic range of approximately 6.02(ENOB−2) dB under its design assumptions.