Pith. sign in

REVIEW 2 major objections 5 minor 58 references

In a full-duplex uplink with movable base-station antennas and a movable-element RIS, secure energy efficiency is maximized by jointly optimizing antenna and RIS positions, phases, powers, receive filters, and artificial noise, and a hybrid

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-01 01:05 UTC pith:U3QPDKVJ

load-bearing objection Useful combination paper with a serious feasibility gap in its main comparison — the H-GML solutions may not satisfy the hard constraints, so the claimed gains over AO are not yet substantiated. the 2 major comments →

arxiv 2607.25924 v1 pith:U3QPDKVJ submitted 2026-07-28 eess.SP

Secure Energy-Efficient Uplink Transmission in Movable-Element RIS-aided Systems with Movable Antennas and Artificial Noise

Ayda Nodel Hokmabadi , Mohamed Elhattab , Chadi Assi This is my paper
classification eess.SP
keywords movable antennasmovable-element RISfull-duplexself-interferencesecure energy efficiencyartificial noisephysical-layer securitygradient-based meta-learning
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper sets out to establish that jointly repositioning both the base station's antennas and the RIS elements, alongside phase shifts, transmit powers, receive filters, and artificial noise, yields higher secure energy efficiency than any fixed-geometry configuration in an uplink with cooperative eavesdroppers. It proposes a hybrid gradient-based meta-learning (H-GML) solver that combines closed-form updates for receive postcoders and the noise direction with learned gradient updates for the remaining coupled variables. Simulation results claim gains of roughly 19% over not using artificial noise, 40% over not knowing eavesdropper channels, about 81% over a fully fixed design without noise, and 15% to 22% over a model-based alternating-optimization benchmark. If correct, the significance is that physical-layer security and energy efficiency can both be improved by reshaping the channel geometry on two sides rather than relying on beamforming alone, and that learned update rules scale more favorably than repeated convex approximation as the surface grows.

Core claim

The central claim is that the secure-energy-efficiency problem in a movable-antenna, movable-element-RIS, full-duplex system with cooperative eavesdroppers can be solved effectively by a hybrid solver: receive postcoders and the artificial-noise direction are updated via generalized Rayleigh quotient closed forms, while user powers, noise power, RIS phases, and the 2D positions of antennas and RIS elements are updated by lightweight neural meta-optimizers trained online from gradients of the SEE objective. The paper reports that this design outperforms all tested baselines, including fixed-geometry, random-RIS, no-AN, no-eavesdropper-CSI, and partial-mobility configurations, with the largest

What carries the argument

The central object is the secure energy efficiency ratio, SEE = (sum of per-user secrecy rates) / (power consumption), together with the field-response channel model, which makes every channel matrix an explicit function of antenna and RIS-element positions through position-dependent phase terms. The workhorse is the hybrid update scheme: a generalized Rayleigh quotient yields the optimal receive postcoder as v_k = R_k^{-1} hbar_k / ||R_k^{-1} hbar_k||, a generalized eigenvalue decomposition gives the AN direction, and six small feed-forward meta-optimizers map normalized gradients of the objective into updates for powers, phases, and positions. Hard constraints are handled by sigmoid power

Load-bearing premise

The paper assumes that the solution returned by Algorithm 1 actually satisfies the hard constraints (QoS, minimum spacing, and movement regions), even though those constraints are only enforced through penalty terms and the algorithm selects the solution with the highest objective value, not the lowest constraint violation.

What would settle it

Run Algorithm 1 on a fixed channel realization, then directly evaluate the returned solution: compute each user's achieved rate and the minimum pairwise distance among all antennas and RIS elements. If any user's rate falls below the QoS threshold or any pair is closer than the minimum spacing d0, the central claim is falsified for that instance. Comparing the SEE of the returned solution against a feasible-repaired version would show how much of the reported gain is due to constraint violation.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • If the reported simulation gains hold, then jointly movable BS antennas and RIS elements offer a concrete mechanism for improving both physical-layer security and energy efficiency in uplink networks.
  • At N = 100 RIS elements, the proposed design claims about 19% higher SEE than the same architecture without artificial noise, and about 40% higher than when eavesdropper CSI is not used, showing that noise injection and Eve awareness are complementary to mobility.
  • The claimed 81% gain over a fully fixed, no-AN baseline suggests that position optimization alone can be a first-order contributor, not a marginal refinement.
  • The H-GML solver is claimed to beat an alternating-optimization benchmark by 15% in energy efficiency and 22% in secure energy efficiency at N = 100, indicating that learned gradient updates can outperform classical block-coordinate methods on this coupled problem.
  • Performance degrades gracefully as the number of cooperative eavesdroppers grows, but the joint mobility plus AN design degrades less than all alternatives.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: If the online meta-learning scheme truly requires no labeled training data, a natural next step is to test it under time-varying channels and user mobility, where retraining a conventional deep network would be prohibitive.
  • Editorial extension: The closed-form AN direction targets the aggregate cooperative-Eve channel and self-interference; a sensible testable extension is a robust variant that assumes imperfect or statistical Eve CSI and compares worst-case secrecy rates.
  • Editorial extension: Because the reported gains rely on penalty-based constraint handling, a deployment-minded extension would add a feasible-repair projection after convergence and measure the resulting SEE loss, if any.
  • Editorial extension: The field-response model assumes far-field plane waves within small movement regions; a near-field extension would likely change how much mobility helps and is a concrete way to stress-test the claimed gains.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. This paper studies secure energy efficiency (SEE) in an uplink full-duplex base station with movable antennas, a movable-element RIS, and multiple cooperative eavesdroppers. It formulates a joint optimization problem (P1) over user powers, receive postcoders, AN power/direction, RIS phases, and antenna/element positions, and proposes a hybrid gradient-based meta-learning (H-GML) solver. H-GML updates receive postcoders and the AN direction via generalized Rayleigh quotient closed forms, and updates the remaining variables with neural-network 'meta-optimizers' trained online on the current channel realization. Simulations claim that H-GML outperforms an AO benchmark by about 15–22% in EE/SEE at N=100 and significantly outperforms fixed-geometry, random-RIS, no-AN, and no-Eve-knowledge baselines.

Significance. If the technical gaps are resolved, the contribution would be significant: it is, to my knowledge, the first joint treatment of dual-sided mobility (MA-BS and ME-RIS) with AN and cooperative Eves under a secrecy-energy-efficiency criterion. The closed-form subproblem updates are standard and correct for their stated objectives, the AO benchmark in Appendix A is carefully specified, and the ablation study covers RIS size, power budgets, AN budget, and Eve count. The complexity analysis is explicit and supports the scalability discussion. The main uncertainties are not in the algebraic derivations but in whether the reported H-GML gains come from feasible solutions and whether the cooperative-Eve SINR model is physically consistent. With feasibility statistics and a corrected Eve model, this would be a solid contribution to the physical-layer-security and movable-antenna literature.

major comments (2)
  1. [§IV-C, Algorithm 1, Fig. 9] The central H-GML-vs-AO comparison is not yet substantiated because H-GML does not enforce the hard constraints C1, C7–C12. In Algorithm 1 (lines 23–25) the returned X* is selected by maximizing the raw objective J=SEE, not by minimizing the penalized losses (27)–(30). The constraints appear only as fixed-weight penalty terms V1, V2, V3 in Eqs. (24)–(26), and there is no projection step or post-hoc feasibility check. The AO benchmark in Appendix A explicitly enforces QoS and spacing constraints (e.g., (55b), (55f)), so an infeasible H-GML solution can trivially achieve higher SEE than a feasible AO solution. The paper must report violation statistics for C1, C7–C12 (e.g., maximum and average violation counts over the 100 realizations), or change the selection criterion/repair step, before the claimed 15–22% gains over AO can be accepted.
  2. [§II-B, Eqs. (2)–(3)] The cooperative-Eve SINR model is not the optimal worst-case combiner. Eq. (3) sums the per-Eve SINRs, which is exact only if the AN-plus-noise terms across the Eves are independent. Here the AN is a common signal observed through distinct scalar channels f_e z, so the AN interference is correlated across Eves. The joint optimal linear combiner gives p_k * h_tilde^H (p_an F z z^H F^H + sigma^2 I)^{-1} h_tilde, not sum_e gamma_{e,k}. Depending on the alignment of h_tilde and F z, the sum formula can overestimate the eavesdropping capability, so the 'worst-case' label is not justified. Because all schemes use the same (incorrect) expression, the relative ranking may be preserved, but the absolute secrecy-rate/SEE values and the physical-layer-security claims are not reliable. Please derive the cooperative SINR from the joint observation model or explicitly justify Eq. (3) as an approximati
minor comments (5)
  1. [Algorithm 1, line 2] Line 2 says 'Apply (22)' before recovering powers; Eq. (22) is the RIS phase retraction. The power recovery should refer to Eqs. (20)–(21).
  2. [Eq. (26)] The penalty V3 uses X0 and rho without definitions. The movement regions U, T, and R should be specified in relation to X0 and rho, otherwise the penalty cannot be evaluated.
  3. [Table II and Section V] The simulation section omits key H-GML hyperparameters: penalty weights mu1, mu2, mu3, hidden width H, network depth, Adam learning rate, Tmax, Ni, No, and the field-response parameters (L, AoA/AoD distributions, path-response coefficients). Without these, the numerical results, especially Fig. 9, cannot be reproduced.
  4. [Eqs. (18)–(19)] When M_t > K, the matrix B in Eq. (19) is singular (rank at most K). The generalized eigenvector update may then be numerically ill-posed. The AO benchmark adds epsilon_z I in Eq. (47); the H-GML update should similarly regularize B with a small identity term.
  5. [Section V, Fig. 9] Fig. 9 appears without a legend or axis labels in the submitted version. Also, with only 100 channel realizations, error bars or confidence intervals would help judge the significance of the reported percentage gains.

Circularity Check

0 steps flagged

No circular derivation: the SEE objective is optimized directly, and the cited self-works are literature context rather than load-bearing premises; the main concerns are feasibility/evaluation design, not circularity.

full rationale

The paper's central claim is that the proposed H-GML solver maximizes the SEE problem (P1) and outperforms an AO benchmark. There is no derivation step in which a target result is assumed as an input. The objective J used in the meta-optimizer losses (27)-(30) is exactly the SEE objective being optimized, and the closed-form postcoder and AN-direction updates (17)-(18) are standard Rayleigh-quotient solutions derived from the same SINR expressions used in the problem. No fitted constant is relabeled as a prediction, and no uniqueness theorem is imported from the authors' prior work to force the design choice. The self-citations [30], [31], [37], [38], [39] are used for literature positioning and system-model motivation, not as the proof of the paper's performance claims. The more substantive issues are non-circular weaknesses: the meta-optimizers are trained online on the same channel realizations and the same SEE objective on which they are evaluated, so the results demonstrate optimization capability rather than out-of-sample generalization; and Algorithm 1 selects X* by the raw objective J (lines 23-25) while constraints (C1), (C7)-(C12) are only penalized through V1-V3, so reported gains over the feasibility-enforcing AO benchmark could partly reflect infeasible solutions. These are correctness and evaluation concerns, not instances of the paper's derivation reducing to its own inputs. Accordingly, no circular step is identified.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 0 invented entities

The central claim rests on modeling assumptions standard in this subfield: far-field channel response, perfect Eve CSI, ideal cooperative Eves, and near-perfect SI cancellation. It also relies on unlisted algorithm design parameters (penalty weights and neural-network hyperparameters) that determine both feasibility and performance. No new physical entities are introduced.

free parameters (4)
  • Penalty weights µ1, µ2, µ3 = not reported
    Balance the objective against QoS, spacing, and movement-region violations in losses (27)-(30); chosen by hand and not specified. They determine whether the returned X* is feasible.
  • Meta-optimizer hyperparameters (hidden width H, depth, Adam learning rate, Tmax, Ni, No) = not reported
    Neural update-rule architecture and training schedule are not given; these directly affect convergence and the reported performance comparison.
  • Field-response simulation parameters (path count L, AoA/AoD distributions, path-response coefficients) = not reported
    Channels (7)-(14) depend on these; Table II lists only basic geometry and path-loss parameters, not L or angular settings.
  • Initialization of powers, phases, antenna/element positions, and random seeds = not reported
    H-GML is a local optimization method; the final solution may depend on initialization, which is not described.
axioms (5)
  • domain assumption Far-field plane-wave approximation: AoAs/AoDs and path gains remain constant within movement regions; only phases vary with position.
    Invoked in Section II-D and Remark 1. The field-response channel model (7)-(14) is built on this; if movement regions are large or near-field effects matter, the position-optimization model is inaccurate.
  • domain assumption The BS has perfect CSI for all links, including all eavesdropper channels, in the main 'with AN' scheme.
    Central to Sections II and V: the proposed design is Eve-aware and the 40% gain over the no-Eve-knowledge baseline depends on this assumption. Passive eavesdroppers generally do not reveal their channels.
  • domain assumption Cooperative Eves share observations over an ideal coordination link and perfectly cancel multiuser interference before decoding.
    Used in Eq. (2) and the text after it. This is a worst-case model; if Eves are less coordinated, the system is more secure, but the design is optimized under this threat model.
  • domain assumption Residual self-interference is modeled by scaling the full SI channel with a small factor η, set to 1e-9 in simulations.
    Section II-B and Table II. This treats SI as nearly perfectly cancelled, which weakens the paper's claim about 'SI suppression' as a main source of gain.
  • standard math Independent noise at the Eves allows the summed-SINR expression (3) to represent maximum-ratio combining.
    Assumed in Section II-B before Eq. (3). The cooperative-eavesdropping rate model is standard but depends on noise independence across Eves.

pith-pipeline@v1.3.0-alltime-deepseek · 24681 in / 14826 out tokens · 149073 ms · 2026-08-01T01:05:23.516825+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Secure Energy-Efficient Uplink Transmission in Movable-Element RIS-aided Systems with Movable Antennas and Artificial Noise." pith.science (2026). https://pith.science/paper/U3QPDKVJ

@misc{pith2026260725924,
  author       = {Pith},
  title        = {Pith review of: Secure Energy-Efficient Uplink Transmission in Movable-Element RIS-aided Systems with Movable Antennas and Artificial Noise},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U3QPDKVJ}},
  note         = {Machine review of arXiv:2607.25924}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Secure energy efficiency (SEE) has emerged as a key performance metric for next-generation wireless networks, where energy sustainability and information security must be jointly guaranteed. This paper investigates secure uplink transmission in a full-duplex (FD) base station (BS) system equipped with movable antennas (MAs) and assisted by a movable-element reconfigurable intelligent surface (ME-RIS) in the presence of multiple cooperative passive eavesdroppers. The objective is to maximize SEE by jointly optimizing the users' transmit powers, BS receive postcoders, artificial noise (AN) transmit power and beamforming, RIS phase shifts, and the two-dimensional positions of both the BS antennas and RIS elements. The resulting optimization problem is highly nonconvex due to the fractional SEE objective, coupled secrecy-rate expressions, residual self-interference (SI), unit-modulus RIS phase-shift constraints, movable-position constraints, inter-element spacing requirements, and the nonlinear dependence of the channels on the movable antenna and RIS-element positions. To address these challenges, we propose a hybrid gradient-based meta-learning (H-GML) framework. In the proposed method, the BS receive postcoders and AN direction are updated using closed-form solutions derived from generalized Rayleigh quotient formulations, while the remaining coupled variables are updated by neural meta-optimizers that learn gradient-based update directions directly from the SEE optimization objective without requiring offline labeled training data. Simulation results show that the proposed H-GML design achieves better performance than the AO benchmark and significantly outperforms fixed-geometry, random RIS, no-AN, and no-Eve-knowledge baselines.

Figures

Figures reproduced from arXiv: 2607.25924 by Ayda Nodel Hokmabadi, Chadi Assi, Mohamed Elhattab.

Figure 1
Figure 1. Figure 1: System model for uplink communications Notation: In this paper, the boldface lowercase and upper￾case letters denote vectors and matrices, respectively. Also, (·) T and (·) H represent the transpose and conjugate transpose. ∥ · ∥ denotes the Euclidean norm. C represents the set of complex numbers. diag(·) is a diagonal matrix from a vector. CN (µ, σ2 ) denotes the circularly symmetric complex Gaus￾sian dis… view at source ↗
Figure 2
Figure 2. Figure 2: Proposed H-GML optimization algorithm structure. [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Convergence behavior of the proposed H-GML framework. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 6
Figure 6. Figure 6: EE/SEE versus the maximum user transmit power [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 5
Figure 5. Figure 5: EE/SEE versus N for different RIS and BS mobility scenarios. N, confirming that joint RIS-element and BS-antenna mobility provides the largest spatial reconfiguration gain. Among the partial-mobility schemes, SEE (FE/MA, with AN) performs better than SEE (ME/FA, with AN), indicating that BS-antenna mobility is particularly effective for enhancing the receive￾side beamforming and SI and eavesdropping suppre… view at source ↗
Figure 8
Figure 8. Figure 8: EE/SEE versus the number of cooperative Eves [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Performance comparison of H-GML and AO versus the N values. [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

58 extracted references · 8 linked inside Pith

  1. [1]

    Communication theory of secrecy systems,

    C. E. Shannon, “Communication theory of secrecy systems,”The Bell system technical journal, vol. 28, no. 4, pp. 656–715, 1949

    1949

  2. [2]

    The wire-tap channel,

    A. D. Wyner, “The wire-tap channel,”Bell system technical journal, vol. 54, no. 8, pp. 1355–1387, 1975

    1975

  3. [3]

    Principles of physical layer security in multiuser wireless networks: A survey,

    A. Mukherjee, S. A. A. Fakoorian, J. Huang, and A. L. Swindlehurst, “Principles of physical layer security in multiuser wireless networks: A survey,”IEEE Communications Surveys & Tutorials, vol. 16, no. 3, pp. 1550–1573, 2014

    2014

  4. [4]

    Modeling and Performance Analysis for Movable Antenna Enabled Wireless Communications,

    L. Zhu, W. Ma, and R. Zhang, “Modeling and Performance Analysis for Movable Antenna Enabled Wireless Communications,”IEEE Trans- actions on Wireless Communications, vol. 23, no. 6, pp. 6234–6250, 2024

    2024

  5. [5]

    A tutorial on movable antennas for wireless networks,

    L. Zhu, W. Ma, W. Mei, Y . Zeng, Q. Wu, B. Ning, Z. Xiao, X. Shao, J. Zhang, and R. Zhang, “A tutorial on movable antennas for wireless networks,”IEEE Communications Surveys & Tutorials, 2025

    2025

  6. [6]

    Movable-Antenna Enhanced Multiuser Communication via Antenna Position Optimization,

    L. Zhu, W. Ma, B. Ning, and R. Zhang, “Movable-Antenna Enhanced Multiuser Communication via Antenna Position Optimization,”IEEE Transactions on Wireless Communications, vol. 23, no. 7, pp. 7214– 7229, 2024

    2024

  7. [7]

    Secure wireless communication via movable-antenna array,

    G. Hu, Q. Wu, K. Xu, J. Si, and N. Al-Dhahir, “Secure wireless communication via movable-antenna array,”IEEE Signal Processing Letters, vol. 31, pp. 516–520, 2024

    2024

  8. [8]

    Enabling Secure Wireless Communications via Movable Antennas,

    Z. Cheng, N. Li, J. Zhu, X. She, C. Ouyang, and P. Chen, “Enabling Secure Wireless Communications via Movable Antennas,” inICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2024, pp. 9186–9190

    2024

  9. [9]

    Movable antenna empowered phys- ical layer security without eve’s CSI: Joint optimization of beamforming and antenna positions,

    Z. Feng, Y . Zhao, K. Yu, and D. Li, “Movable antenna empowered phys- ical layer security without eve’s CSI: Joint optimization of beamforming and antenna positions,”arXiv preprint arXiv:2405.16062, 2024

    Pith/arXiv arXiv 2024

  10. [10]

    Secure MIMO communication relying on movable antennas,

    J. Tang, C. Pan, Y . Zhang, H. Ren, and K. Wang, “Secure MIMO communication relying on movable antennas,”IEEE Transactions on Communications, vol. 73, no. 4, pp. 2159–2175, 2024

    2024

  11. [11]

    Secure full- duplex communication via movable antennas,

    J. Ding, Z. Zhou, C. Wang, W. Li, L. Lin, and B. Jiao, “Secure full- duplex communication via movable antennas,” inGLOBECOM 2024- 2024 IEEE Global Communications Conference. IEEE, 2024, pp. 885– 890

    2024

  12. [12]

    Secure Mul- tiuser Beamforming With Movable Antenna Arrays,

    Z. Cheng, C. Ouyang, B. Zhao, and X. Zhang, “Secure Mul- tiuser Beamforming With Movable Antenna Arrays,”arXiv preprint arXiv:2601.05686, 2026

    Pith/arXiv arXiv 2026

  13. [13]

    Intelligent reflecting surface enhanced wireless network via joint active and passive beamforming,

    Q. Wu and R. Zhang, “Intelligent reflecting surface enhanced wireless network via joint active and passive beamforming,”IEEE transactions on wireless communications, vol. 18, no. 11, pp. 5394–5409, 2019

    2019

  14. [14]

    Wireless communications through reconfigurable intelligent surfaces,

    E. Basar, M. Di Renzo, J. De Rosny, M. Debbah, M.-S. Alouini, and R. Zhang, “Wireless communications through reconfigurable intelligent surfaces,”IEEE access, vol. 7, pp. 116 753–116 773, 2019

    2019

  15. [15]

    Enabling secure wireless commu- nications via intelligent reflecting surfaces,

    X. Yu, D. Xu, and R. Schober, “Enabling secure wireless commu- nications via intelligent reflecting surfaces,” in2019 IEEE Global Communications Conference (GLOBECOM). IEEE, 2019, pp. 1–6

    2019

  16. [16]

    Joint reconfigurable intelligent surface location and passive beamforming optimization for maximizing the secrecy-rate,

    H. Guo, Z. Yang, Y . Zou, B. Lyu, Y . Jiang, and L. Hanzo, “Joint reconfigurable intelligent surface location and passive beamforming optimization for maximizing the secrecy-rate,”IEEE Transactions on Vehicular Technology, vol. 72, no. 2, pp. 2098–2110, 2022

    2098

  17. [17]

    Robust and secure wireless communications via intelligent reflecting surfaces,

    X. Yu, D. Xu, Y . Sun, D. W. K. Ng, and R. Schober, “Robust and secure wireless communications via intelligent reflecting surfaces,”IEEE Journal on Selected Areas in Communications, vol. 38, no. 11, pp. 2637– 2652, 2020

    2020

  18. [18]

    Intelligent reflecting surface: A programmable wireless environment for physical layer security,

    J. Chen, Y .-C. Liang, Y . Pei, and H. Guo, “Intelligent reflecting surface: A programmable wireless environment for physical layer security,”IEEe Access, vol. 7, pp. 82 599–82 612, 2019

    2019

  19. [19]

    Securing reconfig- urable intelligent surface-aided cell-free networks,

    W. Hao, J. Li, G. Sun, M. Zeng, and O. A. Dobre, “Securing reconfig- urable intelligent surface-aided cell-free networks,”IEEE Transactions on Information Forensics and Security, vol. 17, pp. 3720–3733, 2022

    2022

  20. [20]

    Secrecy energy efficiency in RIS-assisted networks,

    Y . Lu, “Secrecy energy efficiency in RIS-assisted networks,”IEEE Transactions on Vehicular Technology, vol. 72, no. 9, pp. 12 419–12 424, 2023

    2023

  21. [21]

    Movable- element RIS-aided wireless communications: An element-wise position optimization approach,

    J. Zhao, Q. Huang, K. Cai, Q. Zhou, X. Mu, and Y . Liu, “Movable- element RIS-aided wireless communications: An element-wise position optimization approach,”IEEE Communications Letters, 2026

    2026

  22. [22]

    RIS-Aided Wireless Communica- tion with Movable Elements: Geometry Impact on Performance,

    Y . Zhang, I. Dey, and N. Marchetti, “RIS-Aided Wireless Communica- tion with Movable Elements: Geometry Impact on Performance,”arXiv preprint arXiv:2405.00141, 2024. 14

    arXiv 2024

  23. [23]

    Movable Antennas Meet Intelligent Reflecting Surface: When Do We Need Movable Antennas?

    X. Wei, W. Mei, Q. Wu, B. Ning, and Z. Chen, “Movable Antennas Meet Intelligent Reflecting Surface: When Do We Need Movable Antennas?” arXiv preprint arXiv:2408.15668, 2024

    Pith/arXiv arXiv 2024

  24. [24]

    Movable Intel- ligent Surface (MIS) for Wireless Communications: Architecture, Mod- eling, Algorithm, and Prototyping,

    H. Li, X. Mu, Y . Liu, K.-K. Wong, and A. Nallanathan, “Movable Intel- ligent Surface (MIS) for Wireless Communications: Architecture, Mod- eling, Algorithm, and Prototyping,”arXiv preprint arXiv:2412.19071, 2024

    Pith/arXiv arXiv 2024

  25. [25]

    Fluid Reconfigurable Intelligent Surface with Element-Level Pattern Reconfigurability: Beamforming and Pattern Co-Design,

    H. Xiao, X. Hu, K.-K. Wong, X. Zhu, H. Hong, and C.-B. Chae, “Fluid Reconfigurable Intelligent Surface with Element-Level Pattern Reconfigurability: Beamforming and Pattern Co-Design,”arXiv preprint arXiv:2508.09695, 2025

    Pith/arXiv arXiv 2025

  26. [26]

    Element Coefficient and Position Optimization for Flexible Intelligent Metasurface-Aided Movable Antenna Communications,

    Q. Bu, S. Yang, J. Ge, B. Ning, and C. Yuen, “Element Coefficient and Position Optimization for Flexible Intelligent Metasurface-Aided Movable Antenna Communications,”IEEE Wireless Communications Letters, 2025

    2025

  27. [27]

    Movable- Element STARS-Aided Secure Communications,

    J. Zhao, Q. Xu, K. Cai, Y . Zhu, X. Mu, and Y . Liu, “Movable- Element STARS-Aided Secure Communications,”IEEE Transactions on Vehicular Technology, 2025

    2025

  28. [28]

    A new design paradigm for secure full-duplex multiuser systems,

    V .-D. Nguyen, H. V . Nguyen, O. A. Dobre, and O.-S. Shin, “A new design paradigm for secure full-duplex multiuser systems,”IEEE Journal on Selected Areas in Communications, vol. 36, no. 7, pp. 1480–1498, 2018

    2018

  29. [29]

    Secure wireless transmission for reconfigurable intelligent surface aided full duplex systems,

    P. Guan, Y . Wang, and Y . Zhao, “Secure wireless transmission for reconfigurable intelligent surface aided full duplex systems,”arXiv preprint arXiv:2206.08797, 2022

    Pith/arXiv arXiv 2022

  30. [30]

    Joint Beamforming and Position Opti- mization for Movable-Antenna and Movable-Element RIS–Aided Full- Duplex 6G MISO Systems,

    A. N. Hokmabadi and C. Assi, “Joint Beamforming and Position Opti- mization for Movable-Antenna and Movable-Element RIS–Aided Full- Duplex 6G MISO Systems,”IEEE Transactions on Communications, 2026

    2026

  31. [31]

    Joint Energy Effi- ciency Optimization for Uplink Multiuser Movable Antenna-Based Wireless Systems Assisted by Movable-Element RIS,

    A. N. Hokmabadi, M. Elhattab, and C. Assi, “Joint Energy Effi- ciency Optimization for Uplink Multiuser Movable Antenna-Based Wireless Systems Assisted by Movable-Element RIS,”arXiv preprint arXiv:2603.28559, 2026

    arXiv 2026

  32. [32]

    Secure Beamforming and Reflection Design for RIS-ISAC Systems under Collusion of Passive and Active Eavesdroppers,

    T. Zhang, Z. Su, and Y . Dong, “Secure Beamforming and Reflection Design for RIS-ISAC Systems under Collusion of Passive and Active Eavesdroppers,”arXiv preprint arXiv:2601.18063, 2026

    Pith/arXiv arXiv 2026

  33. [33]

    Meta-learning based beamforming design for MISO downlink,

    J. Xia and D. Gunduz, “Meta-learning based beamforming design for MISO downlink,” in2021 IEEE International Symposium on Informa- tion Theory (ISIT). IEEE, 2021, pp. 2954–2959

    2021

  34. [34]

    Robust beamforming for RIS- aided communications: Gradient-based manifold meta learning,

    F. Zhu, X. Wang, C. Huang, Z. Yang, X. Chen, A. Al Hammadi, Z. Zhang, C. Yuen, and M. Debbah, “Robust beamforming for RIS- aided communications: Gradient-based manifold meta learning,”IEEE Transactions on Wireless Communications, vol. 23, no. 11, pp. 15 945– 15 956, 2024

    2024

  35. [35]

    Energy-efficient beamforming for RISs-aided communications: Gradient based meta learning,

    X. Wang, F. Zhu, Q. Zhou, Q. Yu, C. Huang, A. Alhammadi, Z. Zhang, C. Yuen, and M. Debbah, “Energy-efficient beamforming for RISs-aided communications: Gradient based meta learning,” inICC 2024-IEEE International Conference on Communications. IEEE, 2024, pp. 3464– 3469

    2024

  36. [36]

    Effi- cient Beamforming Optimization for STAR-RIS-Assisted Communi- cations: A Gradient-Based Meta Learning Approach,

    D. Yang, B. Li, J. He, Y . Yan, X. Zhang, and C. Huang, “Effi- cient Beamforming Optimization for STAR-RIS-Assisted Communi- cations: A Gradient-Based Meta Learning Approach,”arXiv preprint arXiv:2512.17928, 2025

    arXiv 2025

  37. [37]

    Enhancing CoMP-RSMA performance with movable antennas: A meta- learning optimization framework,

    A. Amhaz, S. Khisa, M. Elhattab, C. Assi, and S. Sharafeddine, “Enhancing CoMP-RSMA performance with movable antennas: A meta- learning optimization framework,”IEEE Transactions on Communica- tions, 2026

    2026

  38. [38]

    Meta-Learning Driven Movable-Antenna-assisted Full-Duplex RSMA for Multi-User Communication: Performance and Optimization,

    S. Khisa, A. Amhaz, M. Elhattab, C. Assi, and S. Sharafeddine, “Meta-Learning Driven Movable-Antenna-assisted Full-Duplex RSMA for Multi-User Communication: Performance and Optimization,”arXiv preprint arXiv:2504.03982, 2025

    arXiv 2025

  39. [39]

    Meta-Learning-Driven Resource Optimization in Full-Duplex ISAC with Movable Antennas,

    A. Amhaz, S. Khisa, M. Elhattab, C. Assi, and S. Sharafeddine, “Meta-Learning-Driven Resource Optimization in Full-Duplex ISAC with Movable Antennas,”arXiv preprint arXiv:2510.01437, 2025

    arXiv 2025

  40. [40]

    Meta-reinforcement learning optimization for movable antenna- aided full-duplex CF-DFRC systems with carrier frequency offset,

    Y . Xiu, W. Lyu, Y . Li, R. Yang, P. L. Yeoh, W. Zhang, G. Liu, and N. Wei, “Meta-reinforcement learning optimization for movable antenna- aided full-duplex CF-DFRC systems with carrier frequency offset,”IEEE Transactions on Communications, 2026

    2026

  41. [41]

    Joint beamforming and antenna position optimization for IRS-aided multi- user movable antenna systems,

    Y . Geng, T. H. Cheng, K. Zhong, K. C. Teh, and Q. Wu, “Joint beamforming and antenna position optimization for IRS-aided multi- user movable antenna systems,”IEEE Transactions on Wireless Com- munications, 2025

    2025

  42. [42]

    Movable-antenna aided secure transmission for RIS-ISAC systems,

    Y . Ma, K. Liu, Y . Liu, L. Zhu, and Z. Xiao, “Movable-antenna aided secure transmission for RIS-ISAC systems,”IEEE Transactions on Wireless Communications, 2025

    2025

  43. [43]

    Movable antenna-aided secure full- duplex multi-user communications,

    J. Ding, Z. Zhou, and B. Jiao, “Movable antenna-aided secure full- duplex multi-user communications,”IEEE Transactions on Wireless Communications, 2024

    2024

  44. [44]

    Physical-Layer Security for Full Duplex Communications With Self-Interference Mitigation,

    F. Zhu, F. Gao, T. Zhang, K. Sun, and M. Yao, “Physical-Layer Security for Full Duplex Communications With Self-Interference Mitigation,” IEEE Transactions on Wireless Communications, vol. 15, no. 1, pp. 329– 340, 2016

    2016

  45. [45]

    MIMO capacity characterization for movable antenna systems,

    W. Ma, L. Zhu, and R. Zhang, “MIMO capacity characterization for movable antenna systems,”IEEE Transactions on Wireless Communica- tions, vol. 23, no. 4, pp. 3392–3407, 2023

    2023

  46. [46]

    G. H. Golub and C. F. Van Loan,Matrix computations. JHU press, 2013

    2013

  47. [47]

    Rayleigh quotient based optimization methods for eigenvalue problems,

    R.-C. Li, “Rayleigh quotient based optimization methods for eigenvalue problems,”Matrix Functions and Matrix Equations, vol. 19, pp. 76–108, 2015

    2015

  48. [48]

    On nonlinear fractional programming,

    W. Dinkelbach, “On nonlinear fractional programming,”Management science, vol. 13, no. 7, pp. 492–498, 1967

    1967

  49. [49]

    Absil, R

    P.-A. Absil, R. Mahony, and R. Sepulchre,Optimization algorithms on matrix manifolds. Princeton University Press, 2008

    2008

  50. [50]

    Nocedal and S

    J. Nocedal and S. J. Wright,Numerical optimization. Springer, 2006

    2006

  51. [51]

    Resource allocation for IRS-assisted full-duplex cognitive radio systems,

    D. Xu, X. Yu, Y . Sun, D. W. K. Ng, and R. Schober, “Resource allocation for IRS-assisted full-duplex cognitive radio systems,”IEEE Transactions on Communications, vol. 68, no. 12, pp. 7376–7394, 2020

    2020

  52. [52]

    Boyd and L

    S. Boyd and L. Vandenberghe,Convex optimization. Cambridge university press, 2004

    2004

  53. [53]

    Networked ISAC for low- altitude economy: Transmit beamforming and UA V trajectory design,

    G. Cheng, X. Song, Z. Lyu, and J. Xu, “Networked ISAC for low- altitude economy: Transmit beamforming and UA V trajectory design,” in 2024 IEEE/CIC International Conference on Communications in China (ICCC). IEEE, 2024, pp. 78–83

    2024

  54. [54]

    Joint trajectory and beamforming design for UA V-enabled integrated sensing and communication,

    Z. Lyu, G. Zhu, and J. Xu, “Joint trajectory and beamforming design for UA V-enabled integrated sensing and communication,” inICC 2022- IEEE International Conference on Communications. IEEE, 2022, pp. 1593–1598

    2022

  55. [55]

    Dual-UA V-Aided Covert Communications for Air-to-Ground ISAC Networks,

    J. Sun, L. Yang, A.-A. A. Boulogeorgos, T. A. Tsiftsis, and H. Liu, “Dual-UA V-Aided Covert Communications for Air-to-Ground ISAC Networks,”arXiv preprint arXiv:2506.00601, 2025

    Pith/arXiv arXiv 2025

  56. [56]

    CVX: Matlab software for disciplined convex programming, version 1.21,

    M. Grant, “CVX: Matlab software for disciplined convex programming, version 1.21,”http://cvxr. com/cvx, 2011

    2011

  57. [57]

    Optimization of energy-constrained IRS-NOMA using a complex circle manifold approach,

    M. AlaaEldin, E. Alsusa, K. G. Seddik, M. Al-Jarrah, and C. B. Papadias, “Optimization of energy-constrained IRS-NOMA using a complex circle manifold approach,”IEEE Internet of Things Journal, vol. 11, no. 20, pp. 33 133–33 150, 2024

    2024

  58. [58]

    Minimization of functions having Lipschitz continuous first partial derivatives,

    L. Armijo, “Minimization of functions having Lipschitz continuous first partial derivatives,”Pacific Journal of mathematics, vol. 16, no. 1, pp. 1–3, 1966

    1966