Low-Complexity 6DMA Rotation and Position Optimization Based on Statistical Channel Information

Qijun Jiang; Rui Zhang; Xiaodan Shao

REVIEW 4 major objections 4 minor 22 references

Low-Complexity 6DMA Rotation and Position Optimization Based on Statistical Channel Information

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper proposes and simulates a rotation-first, statistics-based sequential optimizer that configures 6DMA surfaces at a fraction of the computational cost of alternating optimization while keeping comparable average sum log-rate.

desk verdict A plausible low-complexity 6DMA scheme that reduces the problem to a rotation-then-position heuristic, but the performance claim rests on an unquantified Jensen approximation and a restriction mislabeled as a relaxation. read the letter →

arxiv 2507.19309 v1 pith:U7KB7JBT submitted 2025-07-25 cs.IT eess.SPmath.IT

Qijun Jiang , Xiaodan Shao , Rui Zhang This is my paper · ORCID

classification cs.ITeess.SPmath.IT

keywords six-dimensionalmovableantenna6DMAstatisticalchannelinformationsequentialoptimizationalternatingsumlog-ratemaximizationplacementconstraintsgradientascent

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper is trying to show that configuring a six-dimensional movable antenna (6DMA) base station need not alternate between optimizing surface positions and rotations. Its proposed scheme first fixes every surface's rotation while placing the surface center on the largest inscribed sphere of the allowed region, then finds feasible positions that realize those rotations without overlap or mutual blockage. Because the average user rates are replaced by a closed-form lower bound built from channel covariance matrices, the method avoids Monte Carlo averaging over channel realizations. If the paper is right, a low-complexity, statistics-driven procedure can deliver the same average sum log-rate as costly alternating optimization while beating fixed-position and position-only antenna arrays.

What carries the argument

The enabling move is the substitution $z_b(u_b) = [d_{\rm ins} n(u_b)^\top, u_b^\top]^\top$, which ties each surface center to the inscribed sphere of radius $d_{\rm ins}$ inside the feasible volume, so rotation variables alone drive the first-stage objective. The objective is the covariance-based lower bound $\underline{r}_k$ from (19), whose finite-difference gradient is ascended with a backtracking line search; rotations are initialized by a greedy search over a uniformly generated candidate set of rotations. Positions are then found by representing each surface by a circular extended region of radius $d/2$ and sequentially placing tangent planes so that all extended regions lie in the correct half-spaces, guaranteeing the optimized rotations are realizable without surface overlap or mutual blockage.

What would settle it

Take a user-scatterer geometry whose optimal 6DMA configuration is known, by exhaustive search over a coarse grid, to place surface centers near the corners of the feasible volume rather than on the inscribed sphere; run the proposed rotation-first method and the Monte Carlo alternating-optimization benchmark on that geometry and compare average sum log-rate. If the proposed method lags by more than the slack of the Jensen bound, the inscribed-sphere restriction is the cause.

Watch

Extended reading notes

Core claim

The central claim is that the joint 6DMA position-rotation problem can be solved sequentially: optimize rotations first with positions pinned to the inscribed sphere, then repair positions through a geometry-based feasibility algorithm. The rate objective is the analytic Jensen lower bound $\underline{r}_k(\Sigma(z)) = \log_2\left(1+\operatorname{tr}\left(\mathbb{E}[B_k]^{-1}\Sigma_k(z)\right)\right)$, so gradients can be computed by finite differences and no Monte Carlo expectation is needed. The paper reports simulations, with eight four-antenna surfaces, five users, and three dominant scatterers, in which the sequential scheme achieves average sum log-rate close to the Monte Carlo alternating-optimization benchmark at lower complexity, and superior to a fixed three-sector array and a position-adjustable array.

Load-bearing premise

The load-bearing premise is that restricting every surface center to the inscribed sphere during rotation optimization does not exclude the configurations needed for near-optimal rates once positions are later repaired.

Editorial extensions

If this is right

6DMA reconfiguration can be driven by statistical channel information that changes slowly, so mechanical adjustment does not need instantaneous channel knowledge.
The explicit Jensen bound removes the Monte Carlo expectation, cutting the dominant computational cost of the alternating-optimization benchmark while keeping rates comparable.
The geometry-based position repair yields feasible placements inside the 6DMA region, so the optimized rotations are practically implementable rather than abstract.
In the simulated settings, the scheme outperforms fixed-position and position-adjustable baselines, which suggests 6DMA surfaces' rotation freedom, not just position freedom, is what buys the rate gain.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural testable extension is to stress the two-stage ordering in near-field or corner-heavy user geometries, where optimal surface centers may lie far from the inscribed sphere; if the gap to full joint optimization grows there, the position-repair stage would need to feed back into the rotation choice.
The same rotation-first separation could apply to other mechanically steerable surfaces, such as intelligent reflecting surfaces or UAV-mounted arrays, where the slow variable is orientation and position is mostly a feasibility constraint.
The finite-difference gradient over the rotation space is simple but scales with the number of surfaces; an analytic gradient of the Jensen bound would remove the per-surface function-evaluation overhead and is a straightforward extension of this scheme.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

You should know two things. First: this is a genuinely useful direction for 6DMA optimization. The authors replace Monte Carlo alternating optimization with a sequential method: optimize all surface rotations assuming positions lie on the inscribed sphere of the feasible region, then run a geometric CER algorithm to find actual positions that respect blockage and overlap constraints. That decomposition is new relative to their earlier AO work, and using statistical channel information is the right practical framing for mechanically tuned antennas. Second: the headline claim—comparable performance to MC-AO at much lower complexity—is not yet established. The comparison in Fig. 2 appears to pit MC-AO's true Monte Carlo rate (16) against the proposed method's double-Jensen approximation (19), and the paper never states which curve is plotted for the proposed scheme. Two Jensen steps (upper then lower) leave an unquantified gap, so 'comparable' may be partly an artifact.

The paper does good things. The system model is careful, the unified halfspace overlap constraint is sensible, and the CER algorithm is a concrete, well-specified heuristic that does find feasible placements in the simulation. The manuscript is honest in its limitations section insofar as it admits the method is heuristic, though it does not flag the relaxation/restriction problem or the surrogate-rate comparison.

Soft spots, in proportion. The biggest is Eq. (22): the paper calls P2-A a relaxation, but setting q_b = d_ins n(u_b) restricts every surface to the inscribed sphere of V6DMA. That shrinks the feasible set, and no bound or ablation shows the lost position freedom is harmless. Second, the CER algorithm has no proof of success—no guarantee that the greedy shifting yields a feasible configuration, or that the returned positions stay inside V6DMA. Third, the complexity reduction is asserted but never measured; there is no runtime or per-iteration cost comparison. Fourth, the entire performance evaluation is a single simulation with one user/scatterer geometry. None of these are fatal individually, but together they mean the central claim is currently supported by a single, potentially apples-to-oranges curve.

The math and derivations are otherwise sound; the Jensen bounds are correctly applied, and I found no circular reasoning or fitted constants. The self-citation pattern is normal for a group pushing this line of work.

Who is this for? People working on 6DMA or movable-antenna systems will want to read it, and the sequential idea is worth citing even in its current form. It deserves a serious referee: the idea is strong enough that the gaps should be addressed rather than the paper desk-rejected. I would send it out with a request for major revision: quantify the Jensen gap or plot true-rate curves for both schemes, justify or ablate the sphere restriction, and report actual complexity numbers.

For you: worth a skim if you follow the area, but I would not change any of my current work based on the performance claim yet.

Referee Report

4 major / 4 minor

Summary. The paper considers a 6DMA-equipped base station serving K users and optimizes the positions and rotations of B antenna surfaces to maximize the average sum log-rate under a unified blockage/overlap-avoidance constraint. Because Monte Carlo alternating optimization (MC-AO) is expensive, the authors propose a sequential method: first optimize rotations with all surface centers restricted to an inscribed sphere inside the feasible volume (P2-A, Eq. (22)), then construct feasible positions for the optimized rotations via a geometry-based greedy algorithm (P2-B, Eqs. (30)-(42)). The objective in (P1) replaces the true average rate in Eq. (16) by the double-Jensen surrogate in Eq. (19). Section IV reports one simulation comparing the proposed scheme with MC-AO, a fixed-position BS, and a position-adjustable antenna BS, and concludes that the proposed scheme is comparable to MC-AO with much lower complexity.

Significance. If the performance and complexity claims hold, the paper would make 6DMA configuration more practical by replacing Monte Carlo evaluation with a closed-form statistical-channel surrogate and a sequential rotation/position design. Strengths of the manuscript include a parameter-free analytical surrogate (no fitted constants), a greedy initialization strategy for rotations, a geometric position-construction idea, and comparisons against several benchmarks. However, the central claim of 'comparable performance with lower complexity' is not yet established: the surrogate's error relative to the true rate is unquantified, the rotation-only problem is a restriction mislabeled as a relaxation, the position-search heuristic lacks a feasibility proof, and the complexity reduction is not quantified. These are load-bearing gaps, so the manuscript requires major revision.

major comments (4)

[§II-C, Eq. (19) and §IV, Fig. 2] The manuscript never states whether the proposed scheme's curve in Fig. 2 is the true Monte Carlo rate from Eq. (16) or the double-Jensen surrogate from Eq. (19). The MC-AO benchmark explicitly uses W=10^3 Monte Carlo samples to evaluate Eq. (16), while the proposed algorithm optimizes the surrogate in Eq. (19). If the proposed curve is the surrogate, the comparison with MC-AO is not apples-to-apples, and the 'comparable performance' conclusion may be an artifact of the unquantified gap between (16) and (19). Please either evaluate the optimized configuration using Monte Carlo samples of the true rate and plot that curve, or provide a quantitative bound or empirical check showing that the surrogate error is small for the reported configurations.
[§III-A, Eq. (22)] P2-A is called a relaxation of (P1), but setting q_b = d_ins n(u_b) is a restriction: it removes the position variables entirely and confines every surface center to the inscribed sphere of the feasible volume. No argument, proof, or ablation is given that rotations optimizing the restricted problem remain near-optimal after positions are restored in P2-B. If the optimal rotations for the original problem require surface centers that are not on the inscribed sphere, the proposed rotation-first method can be substantially suboptimal. Please provide a dominance or approximation relation between P2-A and P1, or add an ablation that optimizes rotations with free positions (e.g., a coarse position grid) and compares against the inscribed-sphere restriction.
[§III-B, Eqs. (30)-(42)] The geometry-based position search is a heuristic, and the paper does not prove that the returned q satisfies the feasibility constraints (30). In particular, the shift rule in Eq. (40) moves already-positioned surfaces outward along their projected normals; the manuscript does not show that this operation preserves the mutual constraints among those already-positioned surfaces. Moreover, the first surface is assigned 'arbitrarily, for example, the coordinate origin,' which may not lie inside V6DMA in general. Please state and prove the invariant maintained by Steps 1-3, or empirically report the fraction of runs for which all returned positions satisfy (30), and confirm that the simulation's returned positions indeed satisfy the constraints.
[§IV (complexity claim)] The abstract and introduction claim that the proposed scheme 'significantly reduces the computational complexity' of AO, but Section IV provides no runtime, number of objective evaluations, floating-point-operation count, or iteration-count comparison. Since complexity reduction is a central advertised contribution, please quantify it, for example by reporting the average number of objective/gradient evaluations per configuration for the proposed method versus MC-AO, or by reporting wall-clock times.

minor comments (4)

[§II-A, Eq. (8)] Equation (8) appears ill-defined: the antenna-gain matrix G(z, f_{k,l}) depends on the path index l, so it cannot factor out of the concatenated matrix [a(z,f_{k,1}), ..., a(z,f_{k,L_k})] as written. Please write the l-th column as G(z,f_{k,l})a(z,f_{k,l}) and define \tilde A column-wise.
[§II-C and §IV, Fig. 2] The objective in (P1) uses log(r_k) without specifying the base, while Eq. (16) uses log2 and Fig. 2 is labeled 'log-rate.' Please state the base consistently; if natural log is used in (P1), the figure's numerical values will differ from log2 rates by a factor of ln(2).
[§IV, Fig. 2] The simulation reports a single curve per scheme with no confidence intervals, error bars, or description of how many random channel/geometry realizations were averaged. Adding such details would make the 'comparable performance' claim more convincing.
[§III-A, Eqs. (25)-(27)] The gradient is computed by finite differences with epsilon = 2^{-16}; at this scale, cancellation error may be significant for the double-Jensen surrogate. Please comment on the sensitivity of the results to the choice of epsilon or use a complex-step/analytic gradient where possible.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the sequential scheme optimizes an explicit Jensen approximation and the benchmark is independent; concerns are approximation gaps, not self-referential derivations.

full rationale

This paper's derivation is not circular. The objective (P1) maximizes sum log r_k with r_k defined in Eq. (19) as a deterministic function of covariance matrices (20) obtained by two explicit Jensen steps from Eq. (16). This is an approximation with an unquantified gap, but it is not an input-output tautology: there are no fitted constants, no parameter calibrated to a target, and no conclusion imported from the cited works. The rotation stage (P2-A) fixes positions to q_b = d_ins n(u_b) via Eq. (22), which is actually a restriction of the feasible set and is labeled a relaxation; that is a correctness/suboptimality concern, not circularity, because the optimization is still over a well-defined objective and does not presuppose the performance claim. The position stage (P2-B) is a feasibility search, not a refit of the objective. Self-citations in [6]-[9], [11], [16], and [17] supply background, the channel model, and the MC-AO benchmark, but the central sequential decomposition is derived in this paper from the given system model. The main risk—that Fig. 2 may compare the Jensen-approximated rate (19) against MC-AO's Monte Carlo rate (16) without stating that all curves use the true rate—is a statistical-comparison issue, not a circular reduction. Therefore no circular step is identified.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

No physical entity is invented. The cost is paid in model assumptions: known statistical channel information, a double-Jensen surrogate objective, an inscribed-sphere restriction, and an unproven feasibility heuristic. No numbers are fitted to the simulation outcome.

free parameters (1)

Algorithm hyperparameters (tau_ini, delta, kappa_max, epsilon, M) = kappa_max=20, epsilon=2^-16, M=512; tau_ini and delta unspecified
Chosen by hand for the simulations; no sensitivity analysis is reported, so the claimed performance and convergence could depend on these settings.

assumptions (4)

domain assumption Each user's path coefficient vector is complex Gaussian with known diagonal covariance: v_k ~ CN(0, D_k), with D_k known (Eqs. (9)-(11)).
This is the statistical model from which all rate expressions are derived; if the channel is not Rayleigh with independent paths, the covariance-based objective is not the true ergodic rate.
ad hoc to paper The double-Jensen approximation rk in Eq. (19) is treated as the objective to be optimized, with no error bound or Monte Carlo check of its distance from the true rate in Eq. (16).
The paper applies Jensen twice in Eqs. (18) and (19) and then replaces the true expectation by this surrogate; the gap is not characterized.
ad hoc to paper Optimizing rotations with positions fixed on the inscribed sphere in Eq. (22) is assumed to be sufficient for near-optimal performance after the position repair stage.
No proof or ablation shows that this feasible-set restriction does not sacrifice performance; the paper calls it a relaxation but it is a restriction.
ad hoc to paper The geometry-based position search in Section III-B is assumed to return positions satisfying all overlap constraints in Eq. (30).
The algorithm is described stepwise but no feasibility theorem is proved; only one simulation configuration demonstrates it working.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Low-Complexity 6DMA Rotation and Position Optimization Based on Statistical Channel Information." pith.science (2026). https://pith.science/paper/U7KB7JBT

@misc{pith2026250719309,
  author       = {Pith},
  title        = {Pith review of: Low-Complexity 6DMA Rotation and Position Optimization Based on Statistical Channel Information},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U7KB7JBT}},
  note         = {Machine review of arXiv:2507.19309}
}

read the original abstract

The six-dimensional movable antenna (6DMA) is a promising technology to fully exploit spatial variation in wireless channels by allowing flexible adjustment of three-dimensional (3D) positions and rotations of antennas at the transceiver. In this paper, we consider a 6DMA-equipped base station (BS) and aim to maximize the average sum logarithmic rate of all users served by the BS by jointly designing 6DMA surface positions and rotations based on statistical channel information (SCI). Different from prior works on 6DMA design which use alternating optimization to iteratively update surface positions and rotations, we propose a new sequential optimization method that first determines the optimal rotations and then identifies feasible positions to realize these rotations under practical antenna placement constraints. Simulation results show that our proposed optimization scheme significantly reduces the computational complexity of conventional alternating optimization (AO), while achieving communication performance comparable to the AO-based approach and superior to existing fixed-position/rotation antenna arrays.

Figures

Figures reproduced from arXiv: 2507.19309 by the authors.

**Figure 2.** Network average sum log-rate versus transmit power of users. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Illustration of the positions and rotations of 6DMA surfaces optimized [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 10 canonical work pages

[1]

Digital twin-driven network architecture for video streaming,

X. Huang, H. Yang, S. Hu, and X. Shen, “Digital twin-driven network architecture for video streaming,” IEEE Netw., vol. 38, no. 6, pp. 334– 341, 2024

work page 2024
[2]

Beamforming technologies for ultra-massive MIMO in terahertz communications,

B. Ning et al., “Beamforming technologies for ultra-massive MIMO in terahertz communications,” IEEE Open J. Commun. Soc. , vol. 4, pp. 614–658, 2023

work page 2023
[3]

Target sensing with intelligent reflecting surface: Architecture and performance,

X. Shao, C. You, W. Ma, X. Chen, and R. Zhang, “Target sensing with intelligent reflecting surface: Architecture and performance,” IEEE J. Sel. Areas Commun. , vol. 40, no. 7, pp. 2070–2084, Jul. 2022

2022
[4]

Next generation advanced transceiver technologies for 6G and beyond,

C. You et al., “Next generation advanced transceiver technologies for 6G and beyond,” IEEE J. Sel. Areas Commun. , vol. 43, no. 3, pp. 582–627, Mar. 2025

work page 2025
[5]

Intelligent surfaces empowered wireless network: Recent advances and the road to 6G,

Q. Wu et al., “Intelligent surfaces empowered wireless network: Recent advances and the road to 6G,” Proc. IEEE, vol. 112, no. 7, pp. 724–763, Jul. 2024

2024
[6]

6D movable antenna based on user distribution: Modeling and optimization,

X. Shao, Q. Jiang, and R. Zhang, “6D movable antenna based on user distribution: Modeling and optimization,” IEEE Trans. Wireless Commun., vol. 24, no. 1, pp. 355–370, 2025

2025
[7]

6D movable antenna en- hanced wireless network via discrete position and rotation optimization,

X. Shao, R. Zhang, Q. Jiang, and R. Schober, “6D movable antenna en- hanced wireless network via discrete position and rotation optimization,” IEEE J. Sel. Areas Commun. , vol. 43, no. 3, pp. 674–687, 2025

work page 2025
[8]

A tutorial on six-dimensional mov- able antenna for 6G networks: Synergizing positionable and rotatable antennas,

X. Shao, W. Mei, C. You et al. , “A tutorial on six-dimensional mov- able antenna for 6G networks: Synergizing positionable and rotatable antennas,” arXiv preprint arXiv:2503.18240 , 2025

arXiv 2025

Show all 22 references

[9]

Distributed channel estimation and optimization for 6D movable antenna: Unveiling directional sparsity,

X. Shao, R. Zhang, Q. Jiang, J. Park, T. Q. S. Quek, and R. Schober, “Distributed channel estimation and optimization for 6D movable antenna: Unveiling directional sparsity,” IEEE J. Sel. Topics Signal Process., pp. 1–16, 2025

2025
[10]

Polarforming antenna enhanced sensing and communication: Modeling and optimization,

X. Shao, R. Zhang, H. Zhou, Q. Jiang, C. Zhou, W. Zhuang, and X. Shen, “Polarforming antenna enhanced sensing and communication: Modeling and optimization,” arXiv preprint arXiv:2505.08070 , 2025

2025 arXiv
[11]

6DMA-aided cell-free massive MIMO communication,

X. Shi, X. Shao, B. Zheng, and R. Zhang, “6DMA-aided cell-free massive MIMO communication,” IEEE Wireless Commun. Lett., vol. 14, no. 5, pp. 1361–1365, May 2025

2025
[12]

Aerial 6D movable antenna-enabled cell-free networks,

W. Wang et al., “Aerial 6D movable antenna-enabled cell-free networks,” IEEE Trans. Veh. Technol., vol. PP, no. 99, pp. 1–1, 2025

2025
[13]

UA V-enabled passive 6D movable antennas: Joint deployment and beamforming optimization,

C. Liu, W. Mei, P. Wang, Y . Meng, B. Ning, and Z. Chen, “UA V-enabled passive 6D movable antennas: Joint deployment and beamforming optimization,” arXiv preprint arXiv:2412.11150 , 2024

2024 arXiv
[14]

UA V-enabled passive 6DMA for ISAC: Joint location, orientation, and reflection optimization,

P. Wang, Y . Xue, W. Mei, J. Fang, and R. Zhang, “UA V-enabled passive 6DMA for ISAC: Joint location, orientation, and reflection optimization,” arXiv preprint arXiv:2505.10220 , 2025

2025
[15]

Passive six-dimensional movable antenna (6DMA)- assisted multiuser communication,

H. Wang et al. , “Passive six-dimensional movable antenna (6DMA)- assisted multiuser communication,” IEEE Wireless Commun. Lett. , pp. 1–1, early access, 2025

2025
[16]

Hybrid near-far field 6D movable antenna design exploiting directional sparsity and deep learning,

X. Shao et al. , “Hybrid near-far field 6D movable antenna design exploiting directional sparsity and deep learning,” IEEE Trans. Wireless Commun., vol. PP, no. 99, pp. 1–1, 2025

2025
[17]

Exploiting six-dimensional mov- able antenna for wireless sensing,

X. Shao, R. Zhang, and R. Schober, “Exploiting six-dimensional mov- able antenna for wireless sensing,” IEEE Wireless Commun. Lett. , vol. 14, no. 2, pp. 265–269, Feb. 2025

2025
[18]

Performance analysis and optimization for movable antenna aided wideband communications,

L. Zhu, W. Ma, Z. Xiao, and R. Zhang, “Performance analysis and optimization for movable antenna aided wideband communications,” IEEE Trans. Wireless Commun. , vol. 23, no. 12, pp. 18 653–18 668, Dec. 2024

2024
[19]

Movable antenna-aided near-field integrated sensing and communication,

J. Ding et al., “Movable antenna-aided near-field integrated sensing and communication,” arXiv preprint arXiv:2412.19470 , 2024

2024 arXiv
[20]

Nonlinear programming,

D. P. Bertsekas, “Nonlinear programming,” Journal of the Operational Research Society, vol. 48, no. 3, pp. 334–334, 1997

1997
[21]

Technical Specification Group Radio Access Network; Study on 3D Channel Model for LTE,

3rd Generation Partnership Project (3GPP), “Technical Specification Group Radio Access Network; Study on 3D Channel Model for LTE,” 3GPP, Tech. Rep. TR 36.873, 2017

2017
[22]

Particle swarm optimization,

J. Kennedy and R. Eberhart, “Particle swarm optimization,” in Proceed- ings of ICNN’95 - International Conference on Neural Networks , vol. 4, 1995, pp. 1942–1948 vol.4

1995

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Digital twin-driven network architecture for video streaming,

X. Huang, H. Yang, S. Hu, and X. Shen, “Digital twin-driven network architecture for video streaming,” IEEE Netw., vol. 38, no. 6, pp. 334– 341, 2024

work page 2024

[2] [2]

Beamforming technologies for ultra-massive MIMO in terahertz communications,

B. Ning et al., “Beamforming technologies for ultra-massive MIMO in terahertz communications,” IEEE Open J. Commun. Soc. , vol. 4, pp. 614–658, 2023

work page 2023

[3] [3]

Target sensing with intelligent reflecting surface: Architecture and performance,

X. Shao, C. You, W. Ma, X. Chen, and R. Zhang, “Target sensing with intelligent reflecting surface: Architecture and performance,” IEEE J. Sel. Areas Commun. , vol. 40, no. 7, pp. 2070–2084, Jul. 2022

2022

[4] [4]

Next generation advanced transceiver technologies for 6G and beyond,

C. You et al., “Next generation advanced transceiver technologies for 6G and beyond,” IEEE J. Sel. Areas Commun. , vol. 43, no. 3, pp. 582–627, Mar. 2025

work page 2025

[5] [5]

Intelligent surfaces empowered wireless network: Recent advances and the road to 6G,

Q. Wu et al., “Intelligent surfaces empowered wireless network: Recent advances and the road to 6G,” Proc. IEEE, vol. 112, no. 7, pp. 724–763, Jul. 2024

2024

[6] [6]

6D movable antenna based on user distribution: Modeling and optimization,

X. Shao, Q. Jiang, and R. Zhang, “6D movable antenna based on user distribution: Modeling and optimization,” IEEE Trans. Wireless Commun., vol. 24, no. 1, pp. 355–370, 2025

2025

[7] [7]

6D movable antenna en- hanced wireless network via discrete position and rotation optimization,

X. Shao, R. Zhang, Q. Jiang, and R. Schober, “6D movable antenna en- hanced wireless network via discrete position and rotation optimization,” IEEE J. Sel. Areas Commun. , vol. 43, no. 3, pp. 674–687, 2025

work page 2025

[8] [8]

A tutorial on six-dimensional mov- able antenna for 6G networks: Synergizing positionable and rotatable antennas,

X. Shao, W. Mei, C. You et al. , “A tutorial on six-dimensional mov- able antenna for 6G networks: Synergizing positionable and rotatable antennas,” arXiv preprint arXiv:2503.18240 , 2025

arXiv 2025