PaperlayerView exact revision on arXiv

Source-linked AI summary

A Tutorial on Movable Antennas for Wireless Networks

Lipeng Zhu, Wenyan Ma, Weidong Mei, Yong Zeng, Qingqing Wu, Boyu Ning, Zhenyu Xiao, Xiaodan Shao, Jun Zhang, Rui Zhang

arXiv:2502.17905v1cs.ITeess.SP

TL;DR

Fixed antennas limit the spatial degrees of freedom available in varying wireless channels, motivating movable antennas for more adaptive wireless communication and sensing. The paper presents a tutorial covering MA channel models, architectures, optimization, channel acquisition, prototypes, applications, and technology extensions. It reports measured signal-power variation exceeding 40 dB over a 6λ range at 3.5 GHz and about 23 dB at 27.5 GHz, while noting practical movement overhead and unresolved MA–IRS design questions.

  • Problem

    Fixed antennas cannot fully exploit spatial channel variation, limiting spatial degrees of freedom and creating challenges in balancing spectrum efficiency with hardware costs.

  • Method

    The paper develops a comprehensive tutorial covering MA channel models, implementation architectures, optimization, channel acquisition, prototypes, applications, and extensions to other wireless technologies.

  • Results

    Measured signal-power variation exceeded 40 dB over 6λ at 3.5 GHz and was about 23 dB at 27.5 GHz.

  • Takeaways & Limitations

    Movable antennas provide additional spatial degrees of freedom through position and orientation reconfiguration, supporting flexible beamforming and performance enhancement across communication and sensing systems.

Abstract

from arXiv · show

Movable antenna (MA) has been recognized as a promising technology to enhance the performance of wireless communication and sensing by enabling antenna movement. Such a significant paradigm shift from conventional fixed antennas (FAs) to MAs offers tremendous new opportunities towards realizing more versatile, adaptive and efficient next-generation wireless networks such as 6G. In this paper, we provide a comprehensive tutorial on the fundamentals and advancements in the area of MA-empowered wireless networks. First, we overview the historical development and contemporary applications of MA technologies. Next, to characterize the continuous variation in wireless channels with respect to antenna position and/or orientation, we present new field-response channel models tailored for MAs, which are applicable to narrowband and wideband systems as well as far-field and near-field propagation conditions. Subsequently, we review the state-of-the-art architectures for implementing MAs and discuss their practical constraints. A general optimization framework is then formulated to fully exploit the spatial degrees of freedom (DoFs) in antenna movement for performance enhancement in wireless systems. In particular, we delve into two major design issues for MA systems. First, we address the intricate antenna movement optimization problem for various communication and/or sensing systems to maximize the performance gains achievable by MAs. Second, we deal with the challenging channel acquisition issue in MA systems for reconstructing the channel mapping between arbitrary antenna positions inside the transmitter and receiver regions. Moreover, we show existing prototypes developed for MA-aided communication/sensing and the experimental results based on them. Finally, the extension of MA design to other wireless systems and its synergy with other emerging wireless technologies are discussed.

I. INTRODUCTION

Movable antennas (MAs) address the spatial inflexibility of fixed antennas by enabling adaptive antenna placement and geometry. The tutorial surveys their historical development, architectures, channel models, optimization, applications, and practical challenges across wireless networks.

  • A. Background: Fixed antennas cannot fully exploit spatial channel variation, limiting spatial degrees of freedom while antenna costs and energy consumption grow with array size and reconfigurability.This creates a persistent trade-off between spectrum efficiency and hardware costs in wireless networks.
  • A. Background: MAs dynamically change antenna placements and can achieve superior performance with the same or fewer antennas and associated RF chains than fixed-antenna systems.Unlike antenna selection, MAs provide continuous movement within spatial regions rather than only discrete placement choices.
  • B. Historical Development: Antenna movement has a long history, from the Chappe telegraph and Marconi’s kite-supported wire antenna to rotatable directional antennas and formally named movable antennas.The paper also distinguishes fluid antennas as a specific type of MA under their original definition.
  • C. Applications: MA technology is positioned for communication and sensing applications because adaptive movement can improve network adaptability, resilience, efficiency, coverage, and interference management.The surveyed scenarios include IoT, satellite, aerial-platform, and maritime communications.

7) ISAC:

This section frames MA technology as a foundation for adaptive wireless communication and sensing, especially ISAC, by combining flexible antenna movement, generalized channel modeling, hardware architectures, and optimization.

  • ISAC unifies sensing and communication to improve resource utilization and support applications such as autonomous driving and smart cities.
  • The tutorial covers MA architectures, channel models, movement optimization, channel acquisition, prototypes, and extensions to other wireless systems.
  • Architectures: MA hardware architectures are organized by element-level and array-level movement, with their advantages and shortcomings compared for customized system designs.
  • Channel model: The field-response channel model characterizes continuous channel variation with antenna position or orientation across narrowband or wideband, far-field or near-field settings.
  • Optimization: A general optimization framework jointly considers antenna movement and resource allocation to exploit the spatial DoFs of MA systems.
  • Antenna movement model: A 6DMA framework represents flexible antenna position and orientation in 3D space, while lower-DoF systems arise by fixing dimensions or imposing constraints.

2) Extension to MIMO Channel Model:

The MIMO extension defines field-response representations for multiple movable transmit and receive antennas and derives the channel while noting mutual-coupling conditions.

  • For MIMO systems with N_t Tx-MAs and N_r Rx-MAs, field-response matrices represent the responses of all transmit and receive antennas.
  • The resulting MIMO channel matrix is constructed from the transmit and receive field-response matrices and the path response matrix.
  • The model initially neglects mutual coupling, an assumption considered valid when antennas are sufficiently spaced and well isolated.
  • When antennas are closely spaced or poorly isolated, mutual coupling must be modeled through a mutual coupling matrix based on network theory.
  • Wideband extension: For wideband OFDM, clustered delay taps and path-specific transmit and receive wave vectors extend the field-response formulation across the channel impulse response.
  • Wideband extension: The frequency response across all subcarriers is obtained by applying the discrete Fourier transform to the zero-padded channel impulse-response vector.

4) Extension to Near-Field Channel Model:

The tutorial extends field-response channel modeling to near-field propagation and six-dimensional movable antennas, capturing channel variation with antenna positions and orientations. It contrasts this framework with spatial-correlation models and identifies their environmental and radiation-pattern limitations.

  • Near-Field Modeling: Near-field propagation requires spherical-wave modeling when the antenna moving region or carrier frequency places Tx and Rx within the Rayleigh distance.The Tx–Rx distance is represented using reference coordinates, coordinate transforms, and antenna positions.
  • Near-Field Modeling: The near-field channel combines LoS and NLoS components through field responses and a path-response matrix.The model is expressed as hnear(t, r) = hLoS(t, r) + hNLoS(t, r).
  • Model Limitations: Uniform spherical-wave near-field modeling is limited when moving-region size becomes comparable to the Tx–Rx distance, requiring non-uniform spherical waves.The stated near-field model also assumes stationary propagation environments.
  • Six-Dimensional Movable Antennas: The 6DMA model incorporates antenna positions, orientations, radiation patterns, and polarization into a general channel expression.It supports arbitrary omnidirectional or directional radiation patterns and linear, circular, or elliptical polarization.
  • Model Scope: The field-response model explicitly captures continuous channel variation under deterministic propagation environments and supports direct optimization of antenna position and orientation.It can also be applied to stochastic channels, while spatial-correlation models provide simplified approximations in rich-scattering conditions.
  • Model Scope: Spatial-correlation models depend strongly on propagation and fading assumptions and generally cannot characterize antenna-orientation effects on channel gains.Examples include uniform scattering for Jake’s model and Rayleigh fading in many existing works.

C. Fundamental DoFs

Antenna position and orientation provide complementary spatial degrees of freedom for reshaping wireless channels. Their joint optimization can improve channel gains, beamforming flexibility, and sensing performance, with gains depending on propagation paths and antenna patterns.

  • Antenna Position: Changing antenna position primarily adjusts multipath phases, enabling constructive superposition for high channel gain or destructive superposition for low gain.This mechanism applies when antenna orientation is fixed.
  • Antenna Position: Position reconfiguration can provide gains even under pure LoS by changing array geometry and steering vectors in multi-antenna MA arrays.Jointly designing antenna positions and weights supports interference nulling, multi-beamforming, and improved sensing accuracy.
  • Antenna Orientation: Under a single LoS path, orientation optimization can align polarization and radiation patterns, while directional antennas can achieve more orientation-based gain than isotropic antennas.With multiple paths, orientation must balance radiation and polarization gains across paths.
  • Joint Reconfiguration: Joint position and orientation reconfiguration jointly tunes field-response vectors and path-response matrices, providing the highest channel flexibility in 6DMA systems.The available benefit varies as propagation paths and field responses change across scenarios and over time.
  • Case Study: In the Fig. 7 case study, MA/6DMA systems achieve higher channel power gain than FAs, especially as the number of Rx paths increases.For a single path, position optimization provides no gain; orientation can recover 0 dB for isotropic and 6 dB for directional antennas by aligning polarization.

1) Mechanically Movable Element:

MA implementations span mechanically movable, liquid-based, electronically reconfigurable, and array-level architectures. Their differing movement flexibility, speed, range, cost, structural complexity, and control requirements require application-specific selection.

  • Mechanically Movable Element: Mechanically movable elements use actuators such as electric motors, precision gears, or MEMS to position or rotate antennas.Actuators convert control signals and energy into mechanical motion.
  • Liquid-Based Element: Liquid-based elements move fluid or liquid metal through containers using syringes, pumps, or electrowetting control.The mechanism relies on pressure-driven flow or voltage-induced charge redistribution.
  • Electronically Reconfigurable Element: Electronic reconfiguration can emulate antenna movement by shifting a phase center through mode excitation, but it has high hardware cost, complex circuitry, limited displacement, and restricted radiation coverage.These constraints distinguish it from mechanically movable antennas.
  • Array-Level Movement: Array-level architectures include sliding, rotatable, and deployable arrays that alter sub-array spacing, orientation, or overall geometry.Sliding arrays enhance aperture and beam coverage, while rotatable arrays target nonuniform and time-varying user distributions.
  • Practical Constraints: Practical MA design must balance movement flexibility, speed, range, hardware cost, structural complexity, energy consumption, and latency.The appropriate architecture or combination depends on system requirements and antenna scale.

A. MA-Aided SISO Systems

In SISO and array systems, antenna-position optimization reshapes channel phases and array geometry to enhance desired signals, suppress interference, and improve beamforming. The achievable benefit depends on path count, moving-region size, bandwidth, and whether exact steering conditions are attainable.

  • MA-Aided SISO Systems: A single Rx-MA can align multipath phases to maximize channel power gain or destructively combine them to minimize gain.For up to four paths in 3D regions, the stated upper and lower bounds are tight.
  • MA-Aided SISO Systems: As the number of paths increases, channel-power periodicity disappears, but sufficiently large moving regions can still approach the gain bounds.An FPA at the reference point remains limited to one fixed channel power gain.
  • Wideband Systems: Wideband position optimization must balance channel gains across multiple delay taps because one antenna position affects all subcarriers simultaneously.A confined moving region makes it difficult to approach the per-tap bounds across the full channel.
  • Narrowband and Wideband Systems: In Fig. 9, maximum MA gain increases with moving-region size and path count, while wideband gains are smaller than narrowband gains because of frequency selectivity.Narrowband systems can reduce minimum channel power gain by tens of dB relative to FPAs, whereas wideband suppression is weaker.
  • MA Arrays: Jointly optimizing antenna positions and weights enables beam nulling, multi-beamforming, and wide-beam coverage beyond weight-only FPA optimization.The array geometry changes steering vectors and their spatial correlations across directions.
  • MA Arrays: The steering-vector orthogonality condition can enable exact null steering in certain MA configurations, but otherwise only suboptimal solutions may be available.Feasibility depends on the moving-region size, number of antennas, and number of null-steering directions.
  • MA Arrays: An MA array achieved a max-min beamforming gain over 20 dB greater than an FPA array in the reported coverage-region comparison.The MA array also showed reduced beamforming-gain fluctuation within the coverage region.

2) NLoS Channels:

In NLoS MISO channels, MA position optimization is highly nonlinear, so continuous placement can be discretized and solved using graph-based methods. Moderate sampling resolution can approach continuous-search performance while outperforming fixed-antenna benchmarks.

  • NLoS Channels:: The received signal power is PR(x, w) = |h(x)^Hw|^2, with the channel vector determined by the antenna position vector x.For a fixed antenna position, the transmitter can use maximal-ratio transmission to maximize received power.
  • NLoS Channels:: Continuous MA position optimization is difficult because its objective is highly nonlinear with respect to the antenna position vector.Discretizing the moving region converts the problem into a discrete selection problem subject to minimum inter-antenna spacing.
  • NLoS Channels:: The graph-based algorithm models sampling points as graph vertices and uses edge weights based on channel magnitudes to solve the discrete problem optimally in polynomial time.The graph formulation avoids exhaustive search over the combinatorial position-selection space.
  • NLoS Channels:: When M ≥48 or δs ≤λ/6, further sampling-point increases hardly improve received SNR, indicating that moderate sampling resolution is near-optimal.The evaluated setup used Nt = 8 transmit MAs, a 1D moving region of A = 8λ = 0.48 m, and Lt = 7 transmit paths.
  • NLoS Channels:: The optimal graph-based algorithms significantly outperform fixed-position antenna benchmarks as sampling resolution is refined.The comparison includes FPAs with and without antenna selection.

C. MA-Aided MIMO and Multiuser Systems

MA-aided MIMO systems reconfigure channel matrices by jointly optimizing antenna positions and transmission covariance, but the resulting problems are non-convex. Simulations show higher capacity than fixed-position arrays, with statistical CSI approaching instantaneous-CSI performance at high SNR.

  • 1) MIMO Systems:: MA-aided MIMO reshapes channel matrices to improve spatial multiplexing, emphasizing the largest singular value at low SNR and balanced singular values at high SNR.The high-SNR strategy supports water-filling-based spatial multiplexing.
  • 1) MIMO Systems:: Instantaneous-CSI MIMO capacity is optimized jointly over transmit-MA positions, receive-MA positions, and the transmit covariance matrix Q.The covariance matrix satisfies Q ⪰0 and Tr(Q) ≤P.
  • 1) MIMO Systems:: The instantaneous-CSI optimization is challenging because the objective is non-concave and minimum-distance constraints are non-convex.Gradient ascent, successive convex approximation, particle swarm optimization, and alternating optimization are used to obtain suboptimal positions or reduce complexity.
  • 1) MIMO Systems:: Instantaneous-CSI positioning is viable for slowly varying channels, whereas fast-fading systems can incur exorbitant antenna-movement overhead.Statistical-CSI positioning over a longer period can reduce this overhead.
  • 1) MIMO Systems:: MA-MIMO systems always achieve higher capacity than FPA-MIMO systems, while statistical-CSI MA-MIMO approaches instantaneous-CSI performance at high SNR.The results attribute this convergence to positioning being mainly determined by channel-path AoDs and AoAs rather than instantaneous coefficient phases.

2) Multiuser Systems:

In multiuser systems, MA positioning improves spatial multiplexing by enhancing desired channels and reducing inter-user channel correlation. Both instantaneous- and statistical-CSI MA schemes outperform fixed-position benchmarks, while statistical CSI reduces repositioning overhead.

  • 2) Multiuser Systems:: Moving MAs can reduce channel correlation between users, suppress interference power, and significantly improve multiuser spatial multiplexing.The design also exploits constructive signal superposition to increase desired received power.
  • 2) Multiuser Systems:: MA positions can reconfigure user-to-base-station channel vectors and jointly optimize with receive combining and user power allocation.The instantaneous channel vector is h_k(t_k, ˜r) = F_k(˜r)^HΣ_kg(t_k).
  • 2) Multiuser Systems:: Multiuser MA design supports rate-centric optimization of achievable rates under a transmit-power constraint and power-centric optimization of transmit power under a minimum-rate constraint.The objectives can use minimum rate, sum-rate, maximum power, or sum-power metrics.
  • 2) Multiuser Systems:: Non-convex multiuser problems are addressed with position-dependent combining and power allocation together with gradient methods, PSO, OMP, or branch and bound.MMSE and zero-forcing combiners are among the position-dependent choices.
  • 2) Multiuser Systems:: MA schemes always outperform FPA benchmarks in achievable sum rate, and statistical-CSI positioning approaches instantaneous-CSI performance at larger Rician factors.The statistical-CSI scheme offers a balance between communication performance and antenna-repositioning overhead.

D. MA-Aided Sensing and ISAC

MA-aided sensing optimizes antenna positions to improve angle estimation by enlarging effective aperture and reducing CRB. One-dimensional and two-dimensional evaluations report lower MSE than fixed-position array benchmarks, while communication and sensing objectives can conflict.

  • D. MA-Aided Sensing and ISAC: Wireless sensing applications include target distance, speed, and orientation estimation, device localization, and RF imaging.Antenna movement provides additional spatial design freedom for sensing systems.
  • 1) Wireless Sensing:: For 1D AoA estimation, the CRB decreases as the position variance increases, motivating antenna placement that maximizes spatial spread.The minimum-spacing constraint prevents antenna coupling.
  • 1) Wireless Sensing:: The 1D optimal arrangement divides the MAs into two groups at the segment’s leftmost and rightmost ends with minimum inter-antenna spacing.This configuration maximizes array aperture and achieves the best sensing performance.
  • 1) Wireless Sensing:: The 1D optimized MA positions achieve significantly lower AoA-estimation MSE than dense and sparse ULA benchmarks.MUSIC estimates closely approach the CRB at high SNR; the sparse ULA retains grating-lobe problems despite its large aperture.
  • D. MA-Aided Sensing and ISAC: Grating lobes can support multi-beam communication but introduce angular ambiguity and degrade radar sensing accuracy, creating conflicting communication and sensing placement objectives.Balancing these objectives requires deliberate trade-offs.
  • 1) Wireless Sensing:: For 2D sensing, the CRB can be optimized using alternating optimization over the x- and y-coordinate vectors with SCA subproblems.The objective may use maximum CRB or sum-CRB across the two spatial AoAs.
  • 1) Wireless Sensing:: The 2D MA scheme achieves significantly lower AoA-estimation MSE than dense and sparse UPA benchmarks and approaches the CRB lower bound.The MUSIC and CRB curves approach each other in the high-SNR regime.

2) ISAC:

MA-aided ISAC uses movable antenna positions to navigate communication–sensing trade-offs while improving spatial multiplexing, beamforming, and sensing resolution. The section formulates CRB- and capacity-constrained designs and discusses practical deployment challenges.

  • ISAC motivation: MA position reconfiguration can enlarge array aperture, improve sensing angular resolution, boost spatial multiplexing, and reduce steering-vector correlation.These effects support lower angle ambiguity and interference mitigation in communication and sensing.
  • System model: The considered ISAC system estimates a target’s AoA while simultaneously providing communication services using shared hardware and separate time or frequency resources.The sensing metric is the AoA-estimation CRB, while communication performance is represented by capacity or achievable rate.
  • Optimization: The ISAC optimization problems impose either a CRB threshold or a capacity threshold and can use MIMO capacity or multiuser-rate algorithms with SCA relaxation for the sensing metric.The resulting problems are non-convex and are addressed using methods from the corresponding communication optimization designs.
  • Performance trade-off: MA-aided ISAC enables flexible capacity–sensing trade-offs through adaptive antenna positioning, unlike fixed-position benchmark schemes.For sufficiently large 1/ϵs, both instantaneous and statistical MA schemes reconfigure toward the geometry minimizing sensing CRB.
  • Practical considerations: Practical MA-ISAC deployment requires new standardization, infrastructure compatibility, maintenance, reliability, and joint optimization with beamforming and transmit power.These requirements add implementation and design challenges beyond the theoretical position degrees of freedom.

E. MA-Enabled Sparse Array

MA-enabled sparse arrays combine movable geometry with sparsity control to adapt array aperture and spatial resolution across communication, sensing, and ISAC scenarios. The section introduces GMA-based optimization while emphasizing non-convexity and unresolved practical design questions.

  • MA-Enabled Sparse Array: Sparse arrays provide larger aperture, sharper beams, finer spatial resolution, and a larger near-field region than dense arrays with the same element count.Their geometry can also form virtual MIMO through difference or sum co-arrays.
  • MA-Enabled Sparse Array: MA-based sparse arrays can dynamically switch between dense, uniform sparse, and non-uniform sparse geometries according to users, targets, and task requirements.For example, NUSA can support sensing resolution while USA is formed for communication through time switching.
  • GMA architecture: The GMA architecture jointly optimizes array sparsity η through antenna selection and one-dimensional group position x through array movement, reducing movement overhead.The rate objective depends on both x and η under a bounded movement region and maximum sparsity level.
  • Optimization and evaluation: An alternating-optimization algorithm iteratively optimizes x and η for the non-convex sparse-array rate problem.Fig. 25 reports uplink rate versus normalized movable-region size under different ηmax values for a GMA-based sparse array.
  • Open issues: MA-enabled sparse-array research remains early-stage, with joint optimization of antenna movement and general sparse-array geometry still unresolved.Most existing position-optimization studies emphasize narrowband communication setups, and available methods trade optimality against complexity.

1) Robust MA Position Optimization:

Robust MA position optimization addresses imperfect channel knowledge and practical movement constraints that complicate exploiting spatial degrees of freedom. The section also frames channel acquisition as reconstructing the channel over arbitrary transmitter and receiver positions.

  • Robust MA Position Optimization: Perfect instantaneous CSI is difficult to obtain because of channel aging, limited training or feedback, and noise or interference.This motivates position and beamforming designs that explicitly account for imperfect CSI.
  • Robust MA Position Optimization: Deterministic CSI-error models optimize worst-case utility under bounded errors, whereas stochastic models treat CSI errors as random variables, typically Gaussian.Both models provide alternative formulations for robust MA design under uncertainty.
  • Robust MA Position Optimization: Highly nonlinear channel expressions, hardware impairments, and movement-positioning errors make robust MA optimization difficult to solve optimally.These challenges call for more efficient algorithms, especially when imperfect CSI is also present.
  • Movement overhead management: Larger apertures and higher carrier frequencies increase practical concerns about movement implementation, energy consumption, and position-tuning delay.Mechanical drivers add energy use, while movement scheduling remains a largely unresolved challenge.
  • Channel acquisition: MA channel acquisition seeks instantaneous or statistical CSI between arbitrary transmitter and receiver positions to construct a complete channel mapping h(t, r).Exhaustive movement is feasible for small regions but becomes prohibitively costly in training overhead and energy for large regions.
  • Channel acquisition: Model-based acquisition estimates finite field-response parameters, while compressed sensing measures selected locations, estimates wave vectors and PRM, and reconstructs the full mapping.Strategic measurement-location design affects estimation accuracy, creating a direct measurement-planning problem.

2) Tensor Decomposition-based Method:

Tensor decomposition-based acquisition estimates field-response information from structured multi-antenna measurements and uses ESPRIT for grid-free wave-vector recovery. Its super-resolution benefit is limited by antenna-count conditions and UPA-shaped measurement requirements.

  • 2) Tensor Decomposition-based Method: Tensor decomposition-based acquisition estimates Tx-side and Rx-side factor matrices by moving one MA group while fixing the other, then estimates wave vectors with ESPRIT.The procedure uses separate Tx and Rx measurement steps before estimating remaining channel parameters.
  • 2) Tensor Decomposition-based Method: ESPRIT provides grid-free wave-vector estimation, avoiding the discrete wave-vector quantization used by compressed sensing methods.This supports super-resolution FRI estimation in high-SNR regimes.
  • 2) Tensor Decomposition-based Method: The method organizes measurements into a tensor whose factor matrices capture the structured Tx, Rx, and path components of the channel.A UPA-shaped measurement arrangement enables the required decomposition of the channel response.
  • Limitations: The approach requires antenna-count inequalities involving Tx/Rx antennas and paths, making it unsuitable for MA systems with fewer antennas.It also requires UPA-shaped measurement positions, limiting use in irregular moving regions.
  • Method scope: Model-based tensor methods can degrade when the assumed channel model differs from the actual channel, whereas model-free methods avoid channel-model assumptions.Model-free reconstruction instead relies on spatial correlation, Bayesian regression, or machine learning for unmeasured positions.

C. Performance Comparison

Channel acquisition in MA systems involves a trade-off between model-based and model-free methods across region size, SNR, overhead, robustness, and reconstruction accuracy. Experimental prototypes further demonstrate practical sensing and positioning gains from antenna movement.

  • Channel acquisition: Model-based methods offer low measurement overhead and high accuracy in high-dimensional or large regions, but depend on accurate channel structures and are sensitive to measurement errors.Model-free methods are more robust to modeling and measurement errors and suit low-dimensional, small regions, but their overhead grows with region size.
  • Channel acquisition: Above 25 dB SNR, tensor decomposition achieves lower NMSE than compressed sensing through super-resolution FRI estimation using ESPRIT.Compressed-sensing joint FRI estimation outperforms successive estimation above 12 dB, but its computational complexity scales quadratically with successive estimation.
  • Channel acquisition: Channel reconstruction accuracy trades off against time and measurement overhead across model-based and model-free methods as antenna movement-region size changes.The comparison considers a SISO system with an MA moving along a 1D line segment, using 50 channel measurements at 20 dB SNR.
  • Open challenges: Current MA channel-acquisition studies mostly assume fixed antenna orientations and focus on point-to-point communications, leaving rotation and multiuser settings as open challenges.Statistical channel acquisition also remains largely unexplored because long-term instantaneous-channel measurements incur substantial time overhead.
  • Experimental prototypes: MA prototypes have demonstrated more than 50% improved heading-angle estimation, millimeter-order positioning, and decimeter-level positioning accuracy.These results come from mechanical slide-based and motor-driven rotation-type MA systems for localization, sensing, and positioning.

B. Wireless Communication

Experiments and application studies show that movable and reconfigurable antennas can improve wireless communication by exploiting position and orientation changes. Prototypes validate these gains, while broader deployment still requires standardized frameworks, cost-effective architectures, and multi-antenna validation.

  • Prototype validation: A slide-based MA prototype measured over 40 dB signal-power variation across 6λ at 3.5 GHz and about 23 dB at 27.5 GHz.At 3.5 GHz, measured power ranged from −43.7 dBm to −84.1 dBm; at 27.5 GHz, from −84.3 dBm to −107.8 dBm.
  • Prototype validation: Measured and simulated results matched well at both 3.5 GHz and 27.5 GHz, with minor differences attributed to PSI estimation error.The simulations used estimated path state information including path counts, delays, angles of arrival, and path power ratios.
  • Prototype validation: An MA communication system improved spectral efficiency by up to 10% over an FA system in multipath deep fading at 300 GHz.
  • Prototype validation: Fluid and pixel antenna prototypes reduced outage probability by 57%, reached a multiplexing gain of 2.27, and provided around 30 dB signal-power boost at 2.5 GHz.The fluid-antenna result concerns a 4-user system, while the pixel-antenna result validated a 2-user multiple-access system.
  • Lessons learned: Current prototypes mostly use a single antenna, motivating multi-MA validation, cost-effective architectures, and standardized frameworks for practical commercialization.
  • Applications: MA designs extend to satellite links, UAV-mounted systems, AirComp, and MEC through joint movement or placement optimization for interference, computation, and delay objectives.Reported studies reduce interference leakage, mitigate self-interference, reduce computation mean square error, and lower overall delay compared with fixed-antenna systems.

C. Physical Layer Security

Movable antennas improve physical-layer security by adapting antenna positions to strengthen legitimate links and weaken eavesdropper links. The section also identifies unresolved challenges involving imperfect eavesdropper information, broader multiuser MIMO settings, and security risks from malicious use.

  • Secure transmission with movable antennas: Movable antennas can enhance the legitimate channel while degrading eavesdropper links, thereby improving secrecy rates.Position optimization can also reduce channel correlation between legitimate receivers and eavesdroppers in nearby directions.
  • Secure transmission with movable antennas: Joint optimization of artificial-noise beamformers and antenna positions intensifies interference at eavesdroppers and reduces confidential-information leakage.The reviewed studies also report improved covert communication performance using movable antennas.
  • Imperfect eavesdropper information: When eavesdroppers are passive or mobile, statistical CSI and virtual-MA models support secrecy-outage or transmit-power optimization despite unavailable perfect CSI.These approaches model uncertainty by considering potential eavesdropper positions.
  • Open challenges: Existing work has largely studied single-antenna legitimate users and eavesdroppers, leaving MA-aided MIMO security with multiple users and eavesdroppers unexplored.The broader setting requires more sophisticated system modeling and design.
  • Security risks: Movable antennas can also introduce security risks: maliciously used MA-enhanced jamming may dramatically destroy legitimate links, while eavesdroppers may gain enhanced capabilities.The section therefore calls for effective regulations and countermeasures.
Loading 2502.17905v1…