PaperlayerView exact revision on arXiv

Source-linked AI summary

Joint Radar and Communication Design: Applications, State-of-the-art, and the Road Ahead

Fan Liu, Christos Masouros, Athina Petropulu, Hugh Griffiths, Lajos Hanzo

arXiv:1906.00789v1eess.SP

TL;DR

Existing CRSS and DFRC research is concentrated mainly below 6 GHz, motivating mmWave joint radar-communication designs for settings such as V2X. This paper proposes a hybrid analog-digital mmWave DFRC architecture and three-stage TDD frame, with simulations validating joint sensing and communication feasibility.

  • Problem

    Prior research mainly considers sub-6GHz applications, while mmWave DFRC is presented as a newer research area relevant to sensing-enabled communications such as V2X.

  • Method

    The paper proposes a massive-MIMO mmWave DFRC base station with hybrid analog-digital beamforming and a three-stage TDD frame unifying target search, channel estimation, beamforming, and communications.

  • Results

    Simulation results validate the feasibility of realizing radar and communication functionalities on a single mmWave base station, including SIC-based uplink processing.

  • Takeaways & Limitations

    The proposed framework combines target sensing and wireless communication within one mmWave base station while tracking targets and processing uplink signals through SIC.

  • Takeaways & Limitations

    The paper’s scope is constrained by existing research and scenarios that leave broader ranges of constraints and applications for future CRSS work.

Abstract

from arXiv · show

In this paper, we firstly overview the application scenarios and the research progress in the area of communication and radar spectrum sharing (CRSS). We then propose a novel transceiver architecture and frame structure for a dual-functional radar-communication (DFRC) base station (BS) operating in the millimeter wave (mmWave) band, using the hybrid analog-digital (HAD) beamforming technique. We assume that the BS is serving a multi-antenna aided user equipment (UE) operating in a mmWave channel, which in the meantime actively detects multiple targets. Note that part of the targets also play the role of scatterers for the communication signal. Given this framework, we propose a novel scheme for joint target search and communication channel estimation relying on the omni-directional pilot signals generated by the HAD structure. Given a fully-digital communication precoder and a desired radar transmit beampattern, we propose to design the analog and digital precoders under non-convex constant-modulus (CM) and power constraints, such that the BS can formulate narrow beams towards all the targets, while pre-equalizing the impact of the communication channel. Furthermore, we design an HAD receiver that can simultaneously process signals from the UE and echo waves from the targets. By tracking the angular variation of the targets, we show that it is possible to recover the target echoes and mitigate the potential interference imposed on the UE signals by invoking the successive interference cancellation (SIC) technique, even when the radar and communication signals share the equivalent signal-to-noise ratio (SNR). The feasibility and the efficiency of the proposed approaches in realizing DFRC are verified via numerical simulations. Finally, our discussions are summarized by overviewing the open problems in the research field of CRSS.

I. INTRODUCTION

CRSS addresses growing spectrum congestion by coordinating radar and wireless communication, with DFRC jointly integrating communication and sensing. Applications span shared radar-communication bands, vehicular mmWave systems, indoor Wi-Fi sensing, and UAV communication and sensing.

  • Motivation: Spectrum congestion has increased alongside wireless growth, raising the cost of available spectrum and motivating CRSS research.UK operators were required to pay annual totals of £80.3 million for 900 MHz and £119.3 million for 1800 MHz bands since 2015.
  • CRSS directions: CRSS research divides into radar-communication coexistence, which manages mutual interference, and DFRC, which jointly performs communication and remote sensing.DFRC can use a joint system rather than coordinating separate radar and communication platforms.
  • Shared spectrum: Radar and communication systems increasingly share spectrum across L-, S-, C-, and mmWave bands, including cellular, WLAN, GNSS, weather-radar, and automotive-radar uses.The mmWave band is conventionally used for automotive collision-detection and high-resolution imaging radars, while also attracting 5G NR and WLAN communications.
  • Shared spectrum: Interference between base stations and air-traffic-control radars is identified as an urgent coexistence issue for existing and forthcoming networks.The paper notes that airport radar could delay LTE deployment in Southeast England, including major London gateways.
  • Applications: Vehicular V2X networks require low-latency Gbps communications while radar sensing targets centimeter-scale obstacle detection.The paper contrasts tens-of-milliseconds requirements for critical autonomous-vehicle applications with hundreds-of-milliseconds delays tolerated by general communications.
  • Applications: Wi-Fi sensing can support indoor localization and activity recognition by estimating ToA, AoA, RSS, and micro-Doppler-related information from CSI.The paper characterizes this sensing functionality as DFRC and calls for joint signal processing for simultaneous localization and communications.
  • Applications: UAV communication and sensing can support data-demanding services, all-weather radio sensing, formation flight, and collision avoidance, while shared hardware can reduce payload.The dual-functional design aspect for UAVs remains widely unexplored.

C. Military Applications

Military CRSS applications target integrated sensing, communication, electronic warfare, covert communications, and passive radar. The review also identifies limitations of existing coexistence and beamforming approaches, including incomplete spectrum use and radar-detection risks.

  • Multi-function RF systems: Military platforms have historically isolated communication, electronic warfare, and radar subsystems, increasing platform volume, weight, antenna size, and detectability.The paper frames multi-function RF integration as a response to these accumulated hardware costs.
  • Military UAV applications: Military UAV missions such as search and rescue, surveillance, reconnaissance, and electronic countermeasures require both sensing and communication.The integration of these functionalities is presented as a military application of CRSS.
  • Military UAV applications: Communication-enabled UAV monitoring could let cellular base stations detect unauthorized UAVs while continuing to serve authorized UEs without substantial extra hardware.The paper describes cooperative micro base stations in ultra-dense networks as a possible urban air-defense system providing early warning.
  • Radar-assisted LPI communication: Radar-assisted LPI communication can embed communication signals into radar echoes instead of relying solely on frequency/time hopping or spread-spectrum methods.A described RF tag remodulates a probing radar waveform with communication information and returns it through reflected radar signals.
  • Passive radar: Passive radar supports covert operations and avoids extra time/frequency resources, but reliability may be poor because its illumination signal is not tailored or controlled for detection.Joint waveform design and resource allocation are suggested to improve detection probability while maintaining communication performance.
  • RCC limitations: Opportunistic spectrum access limits communication to periods when radar resources are available, while MIMO-radar motion makes sidelobe identification difficult and motivates transmit precoding.The paper states that these approaches do not fully exploit shared spectrum and do not readily extend to MIMO radar.
  • RCC limitations: Null-space-projection approaches cannot exactly control interference power and may zero-force target responses, causing the radar to miss targets.The paper identifies convex optimization as a way to optimize both systems under controllable constraints.

B. Dual-functional Radar-Communication (DFRC) System

DFRC research combines radar sensing and communication through shared waveforms, spatial processing, and joint system design. This paper proposes an mmWave mMIMO architecture and TDD frame that integrate target search, communication-channel estimation, beamforming, tracking, and data transmission.

  • Prior research: Earlier DFRC work explored pulse modulation, OFDM, sidelobe embedding, and spatial processing for carrying communication information while sensing targets.These approaches include PIM, chirp modulation, OFDM range-Doppler processing, ASK or PSK sidelobe signaling, and MIMO beamforming.
  • Motivation: Prior research mainly considered sub-6GHz applications, motivating mmWave DFRC for higher capacity, longer-range sensing, and massive-antenna spatial degrees of freedom.The paper notes that 60GHz WLAN-based sensing uses small arrays and supports only short ranges of tens of meters.
  • Main contributions: The proposed mmWave mMIMO DFRC base station serves a multi-antenna UE while detecting multiple targets, including targets that also scatter the communication signal.Hybrid analog-digital beamforming reduces the required number of RF chains at the transmitter and receiver.
  • Main contributions: A three-stage TDD frame unifies radar target search and channel estimation, radar transmit beamforming and downlink communication, and target tracking and uplink communication.The associated processing includes joint pilots, HAD transmit design, and receiver processing for simultaneous target tracking and uplink decoding.
  • Main contributions: The paper develops joint HAD beamforming that forms directional target beams while equalizing channel effects, plus a receiver that tracks targets and decodes uplink signals.The contribution list explicitly pairs directional beam formation and channel equalization with simultaneous target tracking and uplink decoding.

A. Radar Model

The radar model represents target echoes using transmit probing signals, target reflection coefficients, array steering vectors, and noise-plus-interference. The communication model uses a geometric mmWave channel whose scattering paths correspond to radar-visible targets and give the channel a bistatic-radar interpretation.

  • Radar signal model: The BS transmits a probing matrix over T fast-time snapshots, and the received target echoes depend on target reflection coefficients, azimuth angles, steering vectors, and noise-plus-interference.The model is evaluated in a selected range-Doppler bin, with range and Doppler omitted from the notation.
  • Radar signal model: For a uniform linear array, the steering-vector model uses antenna spacing d set to λ/2.The spacing and wavelength determine the array steering vector.
  • Radar signal model: Arranging target steering vectors into a steering matrix recasts the reflected signal as a matrix model parameterized by target coefficients and azimuth angles.The target coefficient vector contains α1 through αK, while the angle set contains θ1 through θK.
  • Communication signal model: The downlink and uplink communication models describe signals exchanged between the BS and UE through scattering paths, with additive noise at the receiver.The channel is assumed narrowband and constant over duration T, and reciprocity supports the uplink formulation.
  • Radar-communication connection: The mmWave communication channel has intrinsic geometric structure equivalent to a bistatic radar channel, with scatterers acting as known or unknown radar targets depending on channel estimation.This equivalence does not hold for stochastic channel models such as Rayleigh fading, which contain little geometric information.

IV. THE DUAL-FUNCTIONAL RADAR-COMMUNICATION FRAMEWORK

The framework unifies radar target search, communication channel estimation, directional joint beamforming, and target tracking with uplink communication in three coordinated stages.

  • The proposed mmWave DFRC framework unifies radar and communication operations through joint signal processing across three stages.The stages cover target search and channel estimation, directional beamforming, and target tracking with uplink communication.
  • 1) Radar target search and communication channel estimation: Stage 1 uses signals with favorable auto- and cross-correlation properties for joint target search and communication channel estimation.Both radar target extraction and channel characterization require such probing signals before subsequent processing.
  • 2) Joint beamforming: After Stage 1, the BS forms directional beams toward target angles while pre-equalizing communication-channel effects for downlink decoding.The BS uses estimated target angles to obtain more accurate observations and support communication decoding at the UE.
  • 3) Radar target tracking and uplink communication: Stage 3 jointly tracks target-parameter variations and decodes uplink data when target echoes and UE signals are received together.The targets are treated as virtual UEs that reflect probing signals and transmit their geometric parameters noncooperatively.
  • 1) Radar target search and communication channel estimation: Orthogonal LFM pilots generated through time-varying hybrid beamforming produce an omnidirectional beampattern for searching the whole angular domain.The construction recursively updates the analog beamformer and baseband signal while preserving the desired waveform properties.

B. Parameter Estimation

The BS estimates target angles and complex amplitudes from hybrid-combined echoes using a modified MUSIC procedure followed by APES refinement.

  • Random unit-modulus analog combining is used because no prior target AoA information is available during the initial reception stage.The resulting equivalent steering vector is formed after analog combination.
  • A modified MUSIC algorithm estimates target AoAs despite the limited number of massive-MIMO echo observations.The method exploits signal and noise subspaces rather than requiring more receptions than the antenna-array size.
  • The K largest MUSIC-spectrum peaks locate the AoAs of the K targets.The steering vectors at the target angles are orthogonal to the noise subspace.
  • APES estimates each target’s complex amplitude after its angle has been obtained by MUSIC.The method solves an optimization problem for each estimated target angle to improve amplitude accuracy.

SDP ˜YH

The communication-side estimation separates scattering paths from other targets by combining UE angle estimation, zero-forcing uplink pilots, analog combining, and least-squares coefficient estimation.

  • The UE estimates its receive-side angles Φ using MUSIC before forming the uplink beamformer.These estimates support the subsequent zero-forcing transmission of a short uplink pilot.
  • The BS distinguishes communication scatterers from other targets by selecting the largest-magnitude outputs of a targeted analog combiner.The selected outputs identify the AoAs associated with communication paths.
  • The UE’s zero-forcing beamformer removes the steering matrix B(Φ), allowing the BS to observe the communication-path contributions directly.With perfect Φ estimation, the received uplink pilot becomes a target-angle mixture weighted by the scattering coefficients.
  • After identifying the communication paths, the BS estimates their scattering coefficients β_l with a least-squares estimator.The complete procedure is summarized in Algorithm 1.

VI. STAGE 2: RADAR TRANSMIT BEAMFORMING AND DOWNLINK COMMUNICATION

Stage 2 designs a hybrid BS transmit beamformer that simultaneously approximates communication zero-forcing and forms radar beams toward all targets under hybrid-architecture constraints.

  • The proposed joint transmit design forms directional radar beams toward targets while equalizing the downlink communication channel.It uses a hybrid analog-digital structure for both objectives.
  • The transmit signal separates communication-bearing streams from additional radar-only streams because the channel supports only L independent data streams.Both stream groups support radar detection, but only s1 carries downlink communication information.
  • Transmit and receive beamformers jointly equalize the channel because a direct pseudo-inverse is unavailable for the rank-deficient channel matrix.The BS and UE use estimated channel factors to construct corresponding zero-forcing beamformers.
  • B. Low-complexity Approach for DFRC Hybrid Beamforming Design: The analog beamformer is aligned with target steering vectors, but the digital beamformer must also be designed to obtain the desired transmit beampattern.Fixing the analog matrix enables an orthogonal-Procrustes formulation whose global optimum is obtained in closed form.
  • B. Low-complexity Approach for DFRC Hybrid Beamforming Design: The hybrid design scales the precoder to satisfy the total transmit-power budget while approximating the fully digital zero-forcing beamformer.The first L columns preserve the communication design and auxiliary columns support the remaining radar beams.

C. Spectral Efficiency Evaluation

The design forms K narrow beams toward radar targets, and the resulting communication spectral efficiency depends on how the hybrid beamforming manages interference and noise.

  • C. Spectral Efficiency Evaluation: The proposed design guarantees K narrow beams toward the radar targets.When Nt is sufficiently large, the beampattern has peaks only at the target angles.
  • C. Spectral Efficiency Evaluation: The communication beamforming matrix is split into components that separately support communication transmission and account for interference.The post-processing UE signal contains an interference term with covariance matrix Rin.

D. Interference Reduction

The paper reduces interference through hybrid beamforming choices and develops a receiver pipeline for jointly processing overlapped radar echoes and uplink signals.

  • D. Interference Reduction: The choice of Faux is central because interference power is mainly determined by FRF FBB,2, which is optimized to approach Faux.The paper considers both null-space projection and an optimized choice of Faux.
  • D. Interference Reduction: The HBF-Opt design completely eliminates interference by setting Faux = 0 in the hybrid beamforming optimization.Its second digital component lies in the null-space of HFRF, so the associated interference is zero-forced.
  • D. Interference Reduction: The receiver jointly tracks target parameters and decodes uplink data from partially overlapped echo and communication signals.The model separates non-interfered radar, overlapped, and non-interfered uplink portions while using slowly varying angles and newly estimated reflection and scattering coefficients.
  • D. Interference Reduction: The received signal model distinguishes non-overlapped signal components from the mixture occurring during the overlap interval.The radar and uplink signals, together with Gaussian noise, are represented across separate and overlapped periods.
  • D. Interference Reduction: Target-angle variations are assumed small, while target reflection and communication scattering coefficients are re-estimated for each PRI.The angular parameters are tracked from prior estimates, whereas the complex coefficients are treated as independent random realizations.

B. Target Tracking

The target-tracking receiver uses prior angle estimates to form analog receive beams, searches locally for angular changes, reconstructs radar echoes, and then recovers communication signals.

  • B. Target Tracking: The analog combiner maps the received radar and uplink signals into a lower-dimensional RF-chain space.All NRF RF chains are activated, with the first K rows exploiting previously estimated target angles.
  • B. Target Tracking: The first K receive beams point toward previously estimated AoAs, while remaining RF chains provide redundant observations.The redundant observations are intended to improve estimation accuracy.
  • B. Target Tracking: Mutual radar-communication interference does not degrade AoA estimation and may improve estimates for shared target-scatterer angles.Both echoes and communication signals arrive from those angles, increasing their associated signal power.
  • B. Target Tracking: MUSIC searches for AoA peaks within small intervals around prior estimates, bounded by the maximum angular variation.This local search exploits the assumption that angles vary slowly between PRIs.
  • B. Target Tracking: The processing estimates reflection coefficients and angular variations, reconstructs radar echoes, subtracts radar interference, and decodes uplink data using SIC.The receiver subsequently estimates communication path losses and forms a baseband zero-forcing beamformer after analog combination.
  • B. Target Tracking: Synchronization sequences can be orthogonal to the radar signal, but communication data sequences generally are not and therefore still require radar-interference mitigation.The communication signal is recovered only after low-complexity analog combination and radar-signal subtraction.

D. Spectral Efficiency Evaluation

Numerical evaluations show accurate target and channel estimation under favorable conditions, degradation for closely spaced targets at low SNR, and generally improving NMSE with higher SNR and longer pilot duration.

  • D. Spectral Efficiency Evaluation: The simulations use a 64-antenna BS, 16 RF chains, a 10-antenna UE, 8 targets, and 4 communication scatterers.These settings define the main numerical evaluation scenario unless otherwise specified.
  • D. Spectral Efficiency Evaluation: At SNR = 10dB, MUSIC-APES and MUSIC-LS accurately estimate all targets and scatterers, with MUSIC-APES resolving angle pairs separated by 2°.The evaluation uses a 64 × 100 LFM data-pilot matrix and compares estimates with true angles.
  • D. Spectral Efficiency Evaluation: At SNR = 0dB, the BS misses one of two closely spaced targets, while the UE incurs a 1° angle-estimation error.The resulting scattering-coefficient errors are large, but communication performance is reported to be only marginally affected.
  • D. Spectral Efficiency Evaluation: NMSE generally decreases as SNR and data-pilot length increase across radar and communication channel estimation.The results average 8000 random channel realizations while varying both parameters.
  • D. Spectral Efficiency Evaluation: Radar-channel estimation outperforms communication-channel estimation because the BS has more antennas and the uplink pilot is short.The uplink pilot length is fixed at L = 4, which can produce communication-path estimation errors despite good overall communication performance.

B. Radar Transmit Beamforming and Downlink Communication

The proposed DFRC beamforming designs jointly support downlink communication and radar illumination, while tracking targets and mitigating uplink radar-echo interference. Simulations show communication–radar tradeoffs and validate the feasibility of the three-stage mmWave DFRC system.

  • Radar Transmit Beamforming and Downlink Communication: The HBF-Opt design approaches fully digital ZF performance and outperforms HBF-Null under both perfect and estimated CSI.Estimated CSI causes only slight spectral-efficiency losses, supporting the proposed channel-estimation method.
  • Radar Transmit Beamforming and Downlink Communication: The communication-only ZF beamformer forms beams toward four communication scatterers, whereas the proposed DFRC beamformer forms eight beams toward all targets.The DFRC design therefore incorporates targets beyond the communication-channel scatterers.
  • Radar Transmit Beamforming and Downlink Communication: Downlink spectral efficiency decreases as the number of targets increases from K = 8 to 15 at SNR = 20dB.Illuminating more targets diverts power from communication-scatterer directions and reduces SINR; HBF-Opt remains better than HBF-Null.
  • Radar Target Tracking and Uplink Communication: SIC significantly improves uplink spectral efficiency by cancelling radar-echo interference during the overlapped communication period.The gain is marginal for short overlap, weakens when overlap exceeds 90%, and remains considerable in most overlapping cases.
  • Radar Target Tracking and Uplink Communication: At SNR = −20dB, the proposed tracking approach accurately tracks target AoAs at the BS and scattering-path AoDs at the UE with slight errors.The angles are estimated using target echoes and uplink signals at the BS, and communication signals at the UE.
  • Summary of the Proposed Approaches: The three-stage frame structure combines target search and channel estimation, downlink beamforming and communication, and target tracking with uplink communication.The paper reports simulations validating both radar and communication functionality on a single mmWave BS.
  • Future Research Directions: Future work covers broader CRSS constraints and scenarios, including learning-based signal separation, security, V2X channel models, and downlink information-theoretic analysis.The paper identifies these as open research directions rather than resolved capabilities of the presented system.

APPENDIX PROOF OF PROPOSITION 1

The appendix proves Proposition 1 by using the singular value decomposition of F_H and constructing a feasible digital beamforming solution. It verifies unitarity and completes the proof.

  • APPENDIX PROOF OF PROPOSITION 1: The proof introduces the SVD of F_H and separates singular vectors associated with non-zero and zero singular values.The zero-singular-value subspace includes an arbitrary unitary matrix for the remaining dimensions.
  • APPENDIX PROOF OF PROPOSITION 1: The resulting F_BB is verified to be unitary and to satisfy the constraint in problem (41).The proof then applies the constructed expression in a subsequent matrix relation before concluding.
  • APPENDIX PROOF OF PROPOSITION 1: The appendix concludes after establishing the required matrix identities for the proposed solution.The final statement explicitly marks completion of the proof.
Loading 1906.00789v1…