Pith. sign in

REVIEW 4 major objections 4 minor 39 references

The paper argues that two-timescale movable-antenna design, using fast receive-side channel knowledge and slow statistical transmit-side knowledge, significantly raises MIMO downlink sum rates.

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 ↗

A two-timescale movable-antenna design that adapts receive positions to instantaneous channel state information and transmit design to statistical channel state information improves average sum rate and feasibility.

T0 review reviewed 2026-08-05 challenge →

load-bearing objection Useful two-timescale MA-MIMO design with realistic CSI assumptions; the sum-rate results are credible, but the almost-sure KKT claim for the planar-mode algorithm is not actually established and needs either real verification or a softer claim. the 4 major comments →

arxiv 2509.04062 v1 pith:TOCBRIPW submitted 2025-09-04 cs.IT math.IT

Two-Timescale Sum-Rate Maximization for Movable Antenna Enhanced Systems

Xintai Chen , Biqian Feng , Yongpeng Wu , Derrick Wing Kwan Ng , Robert Schober This is my paper
classification cs.IT math.IT
keywords movable antennastwo-timescale optimizationstatistical CSIinstantaneous CSIsum-rate maximizationstochastic successive convex approximationprimal-dual decompositionMIMO downlink
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 reading

The paper tries to show that movable antennas are worthwhile in a realistic multiuser MIMO downlink where the transmitter has only statistical channel knowledge while each receiver knows its own instantaneous channel. It formulates the problem as maximizing average achievable sum rate through a two-timescale decomposition: receive antenna positions adapt quickly to each channel realization, while transmit antenna positions and signal covariance matrices adapt slowly to channel statistics. The central claim is that this split is tractable and profitable—the proposed algorithms outperform fixed uniform planar arrays and movable-antenna systems that rely only on statistics, in both sum rate and feasibility of minimum-rate constraints. A planar movement mode for the receive antennas adds a practical motion-efficiency tradeoff and is claimed to reach stationary (KKT) points almost surely. A sympathetic reader would care because the design avoids the expensive instantaneous CSI feedback that usually limits movable-antenna gains.

Core claim

The central claim is that in a multiuser MIMO downlink with movable antennas, the average achievable sum rate can be meaningfully optimized even when the base station only has statistical channel knowledge, provided each user's receive antennas are re-positioned per coherence interval using their instantaneous channel estimate. The paper decomposes the stochastic problem into short-term receive antenna-position problems, solved by gradient ascent (or gradient projection in the planar mode), and a long-term problem over transmit antenna positions and covariance matrices, solved by stochastic successive convex approximation. For the planar movement mode, the proposed PDD-SSCA variant is claime

What carries the argument

The field-response channel model H_k(t,r_k)=F_k^H(r_k)Σ_k G_k(t), which factors antenna positions through field-response vectors while keeping random path-response matrices, makes position-gradients computable in closed form. On top of it, the two-timescale decomposition: short-term subproblems choose a receive antenna-position vector per I-CSIR sample via gradient ascent or gradient projection, and a long-term CSSCA loop builds convex surrogates of the expected sum-rate objective from averaged gradients and function values. In planar mode, a deep-unrolled gradient recurrence propagates derivatives through the projection step, and the imported PDD-SSCA convergence result supplies the almost-

Load-bearing premise

The almost-sure KKT guarantee for the planar-mode algorithm rests on convergence theorems from reference [38] whose regularity conditions—bounded stochastic gradients, Lipschitz surrogates, and uniform convergence of the short-term gradient-projection solutions—are asserted rather than verified for this problem.

What would settle it

Run Algorithm 4 on a small deterministic instance (e.g., N=2, M=1, K=2) where the projection region boundary is active after many gradient steps, and compute the KKT residual of the limit point while increasing the inner-loop iteration count from small to large. If the residual does not shrink to zero, or if finite-difference checks of the unrolled gradients in equation (33) disagree with the analytical surrogate gradients near a boundary, the O(e^I) omission and the almost-sure KKT claim fail.

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

If this is right

  • Transmitters can avoid the costly instantaneous CSI feedback loop: fast adaptation happens at the receive side, while transmit beams and antenna positions are updated at the statistical time scale.
  • The planar-mode option gives a concrete engineering tradeoff: a small sum-rate loss relative to free movement buys substantially smaller repositioning delay and higher energy efficiency.
  • Minimum-rate constraints become nearly always feasible at low power when receive antennas chase instantaneous channel conditions, so resource allocation is less dominated by worst-case users.
  • Larger antenna movement regions improve average sum rate until saturation, meaning movable-antenna gains scale with the physical space available for motion.
  • If the almost-sure KKT claim holds, the planar-mode algorithm is not just a heuristic: its limit points are certified stationary points of the stochastic problem up to a vanishing error.

Where Pith is reading between the lines

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

  • Editorial inference: the same two-timescale split transfers directly to uplink or cell-free deployments, moving the fast antenna-position updates to the access points and keeping slow statistics at the user side.
  • Editorial inference: the practical throughput depends on physical actuation speed; the paper's delay figures suggest the planar mode, not the general mode, is the regime real motors can sustain over coherence intervals.
  • Editorial inference: a direct test of the deep-unrolled gradient approximation with finite-difference checks at region boundaries would reveal whether the omitted O(e^I) terms are numerically negligible, a question the paper leaves open.
  • Editorial inference: the feasibility-ratio gains suggest antenna movement acts as diversity against fades; quantifying a diversity order from the receive-region size would connect this design to classical diversity-multiplexing reasoning.
Share X Bluesky LinkedIn Reddit HN

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

4 major / 4 minor

Summary. The paper studies a two-timescale resource allocation problem for a multiuser MIMO downlink in which the BS and the UTs are equipped with movable antennas. Long-term transmit antenna positions and transmit covariance matrices are optimized using statistical CSI at the transmitter (S-CSIT), while short-term receive antenna positions are optimized using instantaneous CSI at the receiver (I-CSIR). The authors formulate an average sum-rate maximization with per-user rate constraints and minimum-distance antenna constraints. For the general movement mode (GMM), they decompose the problem into short-term problems solved by gradient ascent and a long-term problem solved by a constrained stochastic successive convex approximation (CSSCA) algorithm. For the planar movement mode (PMM), they propose a primal-dual decomposition-based stochastic successive convex approximation (PDD-SSCA) algorithm, claiming almost-sure convergence to KKT points up to an O(e^{I}) error. Numerical experiments compare the proposed schemes with fixed UPA and S-CSIT-only baselines and report substantial gains in sum rate and feasibility ratio.

Significance. If the KKT guarantee for the PMM were substantiated, the paper would make a meaningful theoretical contribution to MA-enabled systems with two-timescale CSI. The gradient derivations in Appendix A are standard and appear correct; the algorithmic framework is plausible; and the simulation study is broad, covering transmit power, movement-region size, minimum-rate constraints, energy efficiency, and repositioning delay. The optimization pipeline has no fitted parameters, and the reported gains are internally consistent. The main weakness is that the central theoretical claim—almost-sure KKT convergence for the PMM—is not supported by the analysis as written.

major comments (4)
  1. [Sec. IV-A, Eqs. (26)–(28)] Problem (26) contains the expected maximum over the short-term receive APV, cf. Eq. (7). In Sec. IV-A this is replaced by problem (28), in which ~rtilde{r}^{I} is only a stationary point of the short-term problem (27). A KKT statement for (26) requires, by the envelope theorem, that the inner problem be solved to (almost sure) global optimality; a stationary point of a nonconvex problem is not sufficient. The chain-rule terms in (32) do not repair this mismatch, because when the inner point is merely stationary they vanish while the derivative of the max function is evaluated at a different point. Thus Algorithm 4, even under the hypotheses of [38], would at best target (28); the stated KKT guarantee for (26) is not established.
  2. [Sec. IV-C, Eq. (34)] The claim that Algorithm 4 converges to a KKT point of (28) up to an error O(e^{I}) is imported from [38, Theorems 2 and 3] without verifying the theorem hypotheses. Equation (31) only provides pointwise convergence e^{I}(t,Q) → 0; the referenced theorems require additional boundedness/Lipschitz conditions and uniform control of the short-term policy error. Furthermore, the matrices C^{I}_{t,k} and C^{I}_{Q,k,i} in Eq. (34) are needed to be uniformly bounded, but no such bound is proved. The O(e^{I}) bound and the almost-sure claim are therefore unsupported.
  3. [Sec. IV-C, Eq. (33)] The unrolled gradient recursion uses a binary projection mask B^{ℓ}_k. At the boundary of a movement region the projection is non-differentiable, so equality (a) in (33) is not valid in a neighborhood of the boundary. The paper argues that a small enough perturbation in t or {Q_i} cannot change the projection, but no uniform radius is established along the iterative trajectory. This breaks the chain-rule derivation and invalidates the subsequent bound in (34) unless additional regularity is imposed.
  4. [Abstract and Sec. IV-C] The abstract states that PDD-SSCA 'finds KKT solutions almost surely,' omitting the O(e^{I}) approximation error and the limiting condition lim e^{I}=0. The same unqualified wording appears in contribution 3 and in the conclusion. The actual statement near Eq. (34) includes the error, so the claims should be made consistent throughout, with the required hypotheses stated explicitly.
minor comments (4)
  1. [Algorithm 2, input list] The input list includes τ_{r,k}, but this parameter is not used or defined in the algorithm; please remove it or specify its role.
  2. [Sec. IV-C, complexity statement] The complexity analysis of Algorithm 4 omits the cost of computing the unrolled gradients in (32) and (33), which involve matrix derivatives and Hessian-vector products over `I iterations. This cost should be identified and bounded, or explicitly assumed negligible.
  3. [Fig. 3(a)] The caption mentions a dashed line, but the main text and legend do not clearly identify which curve it refers to. Please clarify.
  4. [Eq. (25)] The line 'Since ‖t_i − t_j‖^2 is a convex function ... ∇^2 ‖t_i − t_j‖^2 ⪷ 0' is worded awkwardly; the argument uses the first-order lower bound of a convex function. Please rephrase for clarity.

Circularity Check

0 steps flagged

No significant circularity: the two-timescale sum-rate design is built on direct stochastic-gradient surrogates, and the only imported convergence theorem is external ([38]), not a self-citation.

full rationale

The paper's central derivation chain is not circular. Problem (6) defines the average achievable sum rate directly in (7), with no fitted parameter that is later renamed as a prediction. The GMM short-term receive-APV design (Algorithm 1, Eq. (10)) uses gradients (14) computed from the field-response channel model, and the long-term CSSCA surrogates (18)-(21) use sample gradients (20); these are standard stochastic successive convex approximations, not reductions to the target result. The PMM/PDD-SSCA KKT claim is the only load-bearing theoretical assertion, and it is imported from the external reference [38] ('Based on the proof of [38, Theorems 2 and 3], we can omit these terms...'), not from prior work by the present authors. The paper does not verify all hypotheses of that theorem (e.g., uniform boundedness/Lipschitz conditions for the unrolled gradient in (33)), which is a correctness/rigor concern, but it is not circularity: the theorem is an external result, and the paper's own equations do not make the conclusion equal to an input. The numerical comparisons against UPA and S-CSIT benchmarks are standalone simulations, not fitted predictions. Hence no step in the derivation reduces by construction to its own input.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 0 invented entities

The paper introduces no new physical entities or fitted physical constants. The central claim rests on the far-field field-response channel model, the Gauss-Markov statistical assumption for path gains, and the unverified applicability of two external convergence theorems from [38]. Algorithm hyperparameters in Table I are hand-chosen tuning constants, not fitted to data, but they are listed here for completeness.

free parameters (4)
  • initial step size s = 10
    Hand-chosen in Table I for backtracking line search in Algorithms 1 and 3; affects convergence speed but not the problem definition.
  • backtracking factor τ = 0.5
    Hand-chosen in Table I for the Armijo line search; standard tuning parameter.
  • Armijo parameter ξ = 0.6
    Hand-chosen in Table I to control sufficient increase in the line search.
  • CSSCA step sequences ρ and γ = (ℓ+1)^(-0.9) and (ℓ+1)^(-1)
    Chosen in Table I to satisfy Robbins-Monro conditions in CSSCA; these are algorithmic tuning sequences, not fitted to the data.
axioms (4)
  • domain assumption Far-field channel model: AoDs/AoAs are identical across the transmit and receive regions, so the channel depends on positions only through the phase terms in Eq. (3).
    Section II-A; standard in the MA literature, but limits the model to far-field scattering and planar wavefronts.
  • domain assumption Path gains in Σ_k are i.i.d. zero-mean Gaussian random variables.
    Section II-A; this enables Monte Carlo averaging and a tractable S-CSIT model, but is a specific statistical distribution assumption.
  • domain assumption The conditions of [38, Theorems 2 and 3] hold for the PMM problem (26).
    Section IV-C; the KKT convergence of Algorithm 4 depends on this unverified premise.
  • ad hoc to paper The short-term GP algorithm (Algorithm 3) converges to a stationary point of (27) for every t, Q, H, and the unrolled gradients in (33) are uniformly bounded.
    Section IV-B/C; the O(e^Î) error bound in (34) relies on this, but no proof is given in the paper.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Two-Timescale Sum-Rate Maximization for Movable Antenna Enhanced Systems." pith.science (2026). https://pith.science/paper/TOCBRIPW

@misc{pith2026250904062,
  author       = {Pith},
  title        = {Pith review of: Two-Timescale Sum-Rate Maximization for Movable Antenna Enhanced Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TOCBRIPW}},
  note         = {Machine review of arXiv:2509.04062}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

This paper studies a novel movable antenna (MA)-enhanced multiuser multiple-input multiple-output downlink system designed to improve wireless communication performance. We aim to maximize the average achievable sum rate through two-timescale optimization exploiting instantaneous channel state information at the receiver (I-CSIR) for receive antenna position vector (APV) design and statistical channel state information at the transmitter (S-CSIT) for transmit APV and covariance matrix design. We first decompose the resulting stochastic optimization problem into a series of short-term problems and one long-term problem. Then, a gradient ascent algorithm is proposed to obtain suboptimal receive APVs for the short-term problems for given I-CSIR samples. Based on the output of the gradient ascent algorithm, a series of convex objective/feasibility surrogates for the long-term problem are constructed and solved utilizing the constrained stochastic successive convex approximation (CSSCA) algorithm. Furthermore, we propose a planar movement mode for the receive MAs to facilitate efficient antenna movement and the development of a low-complexity primal-dual decomposition-based stochastic successive convex approximation (PDD-SSCA) algorithm, which finds Karush-Kuhn-Tucker (KKT) solutions almost surely. Our numerical results reveal that, for both the general and the planar movement modes, the proposed two-timescale MA-enhanced system design significantly improves the average achievable sum rate and the feasibility of the formulated problem compared to benchmark schemes.

Figures

Figures reproduced from arXiv: 2509.04062 by Biqian Feng, Derrick Wing Kwan Ng, Robert Schober, Xintai Chen, Yongpeng Wu.

Figure 1
Figure 1. Figure 1: Illustration of the downlink transmission from the BS to [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The proposed PMM architecture (circles represent receive MAs). [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Convergence behavior of the proposed algorithms. [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Average achievable sum rate and feasibility ratio versus the maximum transmit power. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Average energy efficiency and repositioning delay versus the maximum transmit power. [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Average achievable sum rate versus the size of the transmit antenna [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Average achievable sum rate versus the size of the receive antenna [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗
Figure 9
Figure 9. Figure 9: Average achievable sum rate and feasibility ratio versus the required [PITH_FULL_IMAGE:figures/full_fig_p014_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

39 extracted references · 36 canonical work pages · 1 internal anchor

  1. [1]

    Capacity limits of MIMO channels,

    A. Goldsmith, S. A. Jafar, N. Jindal, and S. Vishwanath, “Capacity limits of MIMO channels,” IEEE J. Sel. Areas Commun. , vol. 21, no. 5, pp. 684-702, Jun. 2003

    work page 2003

  2. [2]

    Non-uniform linear antenna array design and optimization for millimeter-wave communica- tions,

    P. Wang, Y . Li, Y . Peng, S. C. Liew, and B. Vucetic, “Non-uniform linear antenna array design and optimization for millimeter-wave communica- tions,” IEEE Trans. Wireless Commun. , vol. 15, no. 11, pp. 7343-7356, Nov. 2016

    work page 2016

  3. [3]

    Non-uniform array design for robust LoS MIMO via convex optimization,

    M. Palaiologos, M. H. C. Garc ´ıa, A. Kakkavas, R. A. Stirling-Gallacher, and G. Caire, “Non-uniform array design for robust LoS MIMO via convex optimization,” Annual International Symposium on Personal, Indoor and Mobile Radio Commun. (PIMRC) , Sep. 2023

    work page 2023

  4. [4]

    Bruce Lee- inspired fluid antenna system: Six research topics and the potentials for 6G,

    K. K. Wong, K. F. Tong, Y . Shen, Y . Chen, and Y . Zhang, “Bruce Lee- inspired fluid antenna system: Six research topics and the potentials for 6G,” Frontiers Commun. Netw., 3:853416, Mar. 2022

    work page 2022

  5. [5]

    Fluid antenna system: New insights on outage probability and diversity gain,

    W. K. New, K. K. Wong, X. Hao, K. F. Tong, and C. B. Chae, “Fluid antenna system: New insights on outage probability and diversity gain,” IEEE Trans. Wireless Commun. , vol. 23, no. 1, pp. 128-140, Jan. 2024

    work page 2024

  6. [6]

    Fluid antenna with linear MMSE channel estimation for large-scale cellular networks,

    C. Skouroumounis and I. Krikidis, “Fluid antenna with linear MMSE channel estimation for large-scale cellular networks,” IEEE Trans. Com- mun., vol. 71, no. 2, pp. 1112-1125, Feb. 2022

    work page 2022

  7. [7]

    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. Wireless Commun., vol. 23, no. 6, pp. 6234-6250, Jun. 2024

    work page 2024

  8. [8]

    MIMO capacity characterization for movable antenna systems,

    W. Ma, L. Zhu, and R. Zhang, “MIMO capacity characterization for movable antenna systems,” IEEE Trans. Wireless Commun. , vol. 23, no. 4, pp. 3392-3407, Apr. 2024

    work page 2024

  9. [9]

    Joint beamforming and antenna movement design for moveable antenna systems based on Statistical CSI,

    X. Chen, B. Feng, Y . Wu, D. W. K. Ng and R. Schober, “Joint beamforming and antenna movement design for moveable antenna systems based on Statistical CSI,” IEEE Global Commun. Conf. , Dec. 2023

    work page 2023

  10. [10]

    Energy efficiency maximiza- tion for movable antenna-enhanced system based on statistical CSI,

    X. Chen, B. Feng, Y . Wu, and W. Zhang, “Energy efficiency maximiza- tion for movable antenna-enhanced system based on statistical CSI,” IEEE Intern. Conf. Commun. , Jun. 2025

    work page 2025

  11. [11]

    Weighted sum-rate maximiza- tion for movable antenna-enhanced wireless networks,

    B. Feng, Y . Wu, X.-G. Xia , and C. Xiao, “Weighted sum-rate maximiza- tion for movable antenna-enhanced wireless networks,” IEEE Wireless Commun. Lett. , vol. 13, no. 6, pp. 1770-1774, Jun. 2024

    work page 2024

  12. [12]

    Movable-antenna array enhanced beam- forming: Achieving full array gain with null steering,

    L. Zhu, W. Ma, and R. Zhang, “Movable-antenna array enhanced beam- forming: Achieving full array gain with null steering,” IEEE Commun. Lett., vol. 27, no. 12, pp. 3340-3344, Oct. 2023

    work page 2023

  13. [13]

    Multi-beam forming with movable- antenna array,

    W. Ma, L. Zhu, and R. Zhang, “Multi-beam forming with movable- antenna array,” IEEE Commun. Lett. , vol. 28, no. 3, pp. 697-701, Jan. 2024

    work page 2024

  14. [14]

    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 Trans. Wireless Commun. , vol. 23, no. 7, pp. 7214-7229, Jul. 2024

    work page 2024

  15. [15]

    Movable antenna-enhanced multiuser communication: Optimal discrete antenna positioning and beamforming,

    Y . Wu, D. Xu, D. W. K. Ng, W. Gerstacker, and R. Schober, “Movable antenna-enhanced multiuser communication: Optimal discrete antenna positioning and beamforming,” IEEE Global Commun. Conf., Dec. 2023

    work page 2023

  16. [16]

    Antenna Positioning and Beamforming Design for Fluid-Antenna Enabled Multi-user Downlink Communications

    H. Qin, W. Chen, Z. Li, Q. Wu, N. Cheng, and F. Chen, “Antenna positioning and beamforming design for movable-antenna enabled multi- user downlink communications,” arXiv preprint arXiv:2311.03046v1, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  17. [17]

    Sum-rate maximization for movable antenna enabled multiuser communications,

    Z. Cheng, N. Li, J. Zhu, X. She, C. Ouyang, and P. Chen, “Sum-rate maximization for movable antenna enabled multiuser communications,” arXiv preprint arXiv:2309.11135v1, 2023

    Pith/arXiv arXiv 2023

  18. [18]

    Multiuser commu- nications with movable-antenna base station: Joint antenna positioning, receive combining, and power control,

    Z. Xiao, X. Pi, L. Zhu, X. G. Xia, and R. Zhang, “Multiuser commu- nications with movable-antenna base station: Joint antenna positioning, receive combining, and power control,” IEEE Trans. Wireless Commun., vol. 23, no. 12, pp. 19744-19759, Dec. 2024

    work page 2024

  19. [19]

    Capacity maximization for FAS-assisted multiple access channels,

    H. Xu, K. K. Wong, W. K. New, G. Zhou, R. Murch, C. B. Chae, Y . Zhu, and S. Jin, “Capacity maximization for FAS-assisted multiple access channels,” IEEE Trans. Commun. , early access, Dec. 2024

    work page 2024

  20. [20]

    Movable antenna-empowered AirComp,

    Z. Cheng, N. Li, J. Zhu, X. She, C. Ouyang, and P. Chen, “Movable antenna-empowered AirComp,” arXiv preprint arXiv:2309.12596v1, 2023

    arXiv 2023

  21. [21]

    Movable- antenna array-enabled wireless communication with CoMP reception,

    G. Hu, Q. Wu, J. Ouyang, K. Xu, Y . Cai, and N. A. Dhahir, “Movable- antenna array-enabled wireless communication with CoMP reception,” IEEE Commun. Lett. , vol. 28, no. 4, pp. 947-951, Apr. 2024

    work page 2024

  22. [22]

    Secure wireless communication via movable-antenna array,

    G. Hu, Q. Wu, K. Xu, J. Si, and N. A. Dhahir, “Secure wireless communication via movable-antenna array,” IEEE Signal Process. Lett. , vol. 31, pp. 516-520, Jan. 2024

    work page 2024

  23. [23]

    Compressed sensing based channel estimation for movable antenna communications,

    W. Ma, L. Zhu, and R. Zhang, “Compressed sensing based channel estimation for movable antenna communications,” IEEE Commun. Lett. , vol. 27, no. 10, pp. 2747-2751, Oct. 2023

    work page 2023

  24. [24]

    Channel estimation for movable antenna communication systems: A framework based on compressed sensing,

    Z. Xiao, S. Cao, L. Zhu, Y . Liu, X. G. Xia, and R. Zhang, “Channel estimation for movable antenna communication systems: A framework based on compressed sensing,” IEEE Trans. Wireless Commun. , vol. 23, no. 9, pp. 11814 - 11830, Sep. 2024

    work page 2024

  25. [25]

    Completely stale transmitter channel state information is still very useful,

    M. A. Maddah-Ali and D. Tse, “Completely stale transmitter channel state information is still very useful,” IEEE Trans. Inf. Theory , vol. 58, no. 7, pp. 4418–4431, Jul. 2012

    work page 2012

  26. [26]

    MIMO broadcast channels with finite-rate feedback,

    N. Jindal, “MIMO broadcast channels with finite-rate feedback,” IEEE Trans. Info. Theory , vol. 52, no. 11, pp. 5045–5060, Nov. 2006

    work page 2006

  27. [27]

    FDD massive MIMO via UL/DL channel covariance extrapolation and active channel sparsification,

    M. Barzegar Khalilsarai, S. Haghighatshoar, X. Yi, and G. Caire, “FDD massive MIMO via UL/DL channel covariance extrapolation and active channel sparsification,” IEEE Trans. Wireless Commun. , vol. 18, no. 1, pp. 121–135, Jan. 2019

    work page 2019

  28. [28]

    K. B. Petersen and M. S. Pedersen, The Matrix Cookbook . Kgs. Lyngby, Denmark: Tech. Univ. Denmark, 2006

    work page 2006

  29. [29]

    Antenna arrays for line-of-sight massive MIMO: Half wavelength is not enough,

    D. Pinchera, M. D. Migliore, F. Schettino, and G. Panariello, “Antenna arrays for line-of-sight massive MIMO: Half wavelength is not enough,” Electronics, vol. 6, no. 3, pp. 57-57, Sep. 2017

    work page 2017

  30. [30]

    Fluid antenna- assisted MIMO transmission exploiting statistical CSI,

    Y . Ye, L. You, J. Wang, H. Xu, K. K. Wong, and X. Gao, “Fluid antenna- assisted MIMO transmission exploiting statistical CSI,” IEEE Commun. Lett., vol. 28, no. 1, pp. 223-227, Nov. 2023

    work page 2023

  31. [31]

    Stochastic successive convex approximation for nonconvex constrained stochastic optimization,

    A. Liu, V . K. N. Lau, and B. Kananian, “Stochastic successive convex approximation for nonconvex constrained stochastic optimization,” IEEE Trans. Signal Process. , vol. 67, no. 16, pp. 4189-4203, Aug. 2019

    work page 2019

  32. [32]

    J. R. Magnus and H. Neudecker, Matrix Differential Calculus With Applications in Statistics and Econometrics . Hoboken, NJ, USA: Wiley, 1995

    work page 1995

  33. [33]

    E. K. Chong and S. H. Zak, An Introduction to Optimization . Hoboken, New Jersey, USA: John Wiley & Sons, 2013

    work page 2013

  34. [34]

    Boyd and L

    S. Boyd and L. Vandenberghe, Convex Optimization. Cambridge, U.K.: Cambridge Univ. Press, 2004

    2004

  35. [35]

    Cottle, Nonlinear Programming

    R. Cottle, Nonlinear Programming. Hoboken, New Jersey, USA: John Wiley & Sons, 1972

    work page 1972

  36. [36]

    Flexible-position MIMO for wireless communications: fundamentals, challenges, and future directions,

    J. Zheng, J. Zhang, H. Du, D. Niyato, S. Sun, Bo Ai, and K. B. Letaief, “Flexible-position MIMO for wireless communications: fundamentals, challenges, and future directions,” IEEE Wireless Commun., vol. 31, no. 5, pp. 18-26, Oct. 2023

    work page 2023

  37. [37]

    Majorization-minimization algo- rithms in signal processing, communications, and machine learning,

    Y . Sun, P. Babu, and D. P. Palomar, “Majorization-minimization algo- rithms in signal processing, communications, and machine learning,” IEEE Trans. Signal Process. , vol. 65, no. 3, pp. 794-816, Feb. 2017

    work page 2017

  38. [38]

    Two-stage stochastic optimization via primal-dual decomposition and deep unrolling,

    A. Liu, R. Yang, T. Q. S. Quek, and M. Zhao,“Two-stage stochastic optimization via primal-dual decomposition and deep unrolling,” IEEE Trans. Signal Process. , vol. 69, pp. 3000-3015, 2021

    work page 2021

  39. [39]

    Complex-valued matrix differentiation: Techniques and key results,

    A. Hjørungnes, and D. Gesbert, “Complex-valued matrix differentiation: Techniques and key results,” IEEE Trans. Signal Process. , vol. 55, no. 6, pp. 2740 - 2746, Jun. 2007

    work page 2007

This paper was first reviewed by deepseek-v4-flash on August 5, 2026.