Simultaneous Beamforming and Anti-Jamming With Intelligent Omni-Surfaces

Boya Di; Hongliang Zhang; Qingyu Liu; Shuhao Zeng; Yuhan Wang

REVIEW 3 major objections 5 minor 15 references

Simultaneous Beamforming and Anti-Jamming With Intelligent Omni-Surfaces

T0 review · 3 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read An intelligent omni-surface that reflects and refracts simultaneously can nullify jamming while enhancing desired signals, outperforming IRS-based anti-jamming in multiuser downlink systems.

desk verdict The central SDP relaxation is invalid because constraint (23e) is nonconvex, so the algorithm as written cannot deliver the claimed results; the problem setup is still worth a serious look. read the letter →

arxiv 2502.02055 v1 pith:XH2KS4FT submitted 2025-02-04 eess.SP

Yuhan Wang , Shuhao Zeng , Qingyu Liu , Boya Di , Hongliang Zhang This is my paper · ORCID

classification eess.SP

keywords intelligentomni-surfaceanti-jammingreflectionandrefractionbeamformingsemidefiniterelaxationzero-forcingphaseshiftdesignmultiuserdownlink

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 claims that a single intelligent omni-surface (IOS) can both nullify jamming and enhance desired signals at the same time, something reflective-only intelligent reflecting surfaces (IRS) cannot fully achieve because they serve only one side of the surface. The authors propose to jointly design the base station's digital beamformer and the IOS's coupled reflection and refraction phase shifts, despite the phases being discrete and the jammer's beamformer and channel state information being imperfect. They relax the discrete coupled phases to continuous ones, impose a linear coupling relation between them, solve the resulting semidefinite program using the Cauchy-Schwarz inequality and the S-procedure, then recover discrete states by a local search. Simulation shows the scheme raises sum rate above IRS-based enhancement-only or cancellation-only schemes and brings 94% of users' received jamming power below -96 dBm. If the scheme holds, an IOS becomes a practical single-device answer to jamming in multi-user systems.

What carries the argument

The load-bearing object is the IOS element model with a finite set $\mathcal{F}$ of coupled reflection/refraction phase pairs $(\phi^r_m, \phi^t_m)$, together with the linear coupling constraint $\phi^t_m = s\phi^r_m + v$ (equation 21b) that lets both phase vectors be encoded in a single rank-one matrix $\Xi_r$. The algorithm then uses the Cauchy-Schwarz inequality to bound and eliminate the unknown jammer beamformer, the S-procedure to absorb bounded channel uncertainties, and semidefinite relaxation followed by Gaussian randomization to obtain continuous phases, after which a one-element-at-a-time local search over the two bracketing discrete pairs in $\mathcal{F}$ recovers feasible discrete states.

What would settle it

Take a fabricated IOS element, measure its actual set of $2^b$ achievable pairs $(\phi^r, \phi^t)$, and check how far they deviate from the best linear fit. If many pairs have residuals comparable to the quantization step, or if the fitted line excludes pairs that would be needed to satisfy the jamming constraint, then the continuous relaxation in Step 1 and the local search in Step 3 would miss feasible states and the claimed jamming nulling and sum-rate gains would not hold in practice. A direct numerical falsifier is to re-run the proposed algorithm with a measured (non-collinear) phase-pair set and compare the achieved sum rate and jamming-violation probability against the paper's Figure 2 curves.

Watch

Extended reading notes

Core claim

The central claim is that an IOS can simultaneously cancel jamming and enhance desired signals by applying different phase shifts to the reflected and refracted components of the impinging wave, even when those phase shifts are discrete and coupled per element. The paper formulates a sum-rate maximization problem under a per-user jamming-power constraint, decomposes it into a digital zero-forcing beamforming subproblem solved by water-filling and an analog IOS phase-shift subproblem. To handle the coupled discrete phases, the authors fit a linear relation between reflection and refraction phases, relax the problem to a continuous semidefinite program in one rank-one matrix that encodes both phase vectors, eliminate the unknown jammer beamformer with a Cauchy-Schwarz bound, and handle channel uncertainty with the S-procedure. A local search over the feasible discrete phase-pair set then recovers discrete states. In 28 GHz simulations, the IOS scheme yields a higher sum rate than IRS-assisted signal-enhancement or jamming-cancellation schemes, and a 1-bit IOS already outperforms a high-bit IRS, demonstrating that simultaneous reflection and refraction, not fine phase quantization, drives the gain.

Load-bearing premise

The whole scheme relies on the assumption that the discrete reflection and refraction phase pairs available on a real IOS element lie on a single straight line, $\phi^t_m = s\phi^r_m + v$, and that a linear fit captures this relation well enough; the paper uses this line in the relaxation and in the final local search, but provides no fitting procedure, no fitted parameter values, and no error measure.

Editorial extensions

If this is right

A single IOS deployed between a base station and a jammer can protect users on both sides of the surface, removing the need for two separate IRSs for signal enhancement and jamming cancellation.
The numerical result that a 1-bit IOS outperforms a high-bit IRS suggests that the dual-sided reflection/refraction capability matters more than fine phase quantization for anti-jamming.
The coupling-aware SDP formulation makes the joint design tractable: it reduces a non-convex discrete coupled problem to a convex relaxation plus a low-complexity local search.
If the linear coupling fit is accurate, the same algorithmic pipeline applies to any metasurface whose reflection and refraction phases are related by a fixed curve, not just IOS.
The jamming-nulling performance (94% of users below noise floor -96 dBm) suggests the scheme can push residual jamming below thermal noise in typical 28 GHz deployments.

Reading between the lines

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

The paper's dependence on the linear coupling constraint (21b) is both its enabler and its fragile point: in a real device the measured phase pairs may not be collinear, and the paper does not report the fitting procedure, the fitted values of $s$ and $v$, or the fitting error, so the claimed gains rest on an unverified premise about hardware behavior.
A direct testable extension would be to take a measured set of $2^b$ phase pairs from an IOS prototype, fit the line, and re-run the proposed algorithm; if the sum-rate gap over IRS narrows or the jamming constraint is violated, the hardware coupling model needs revision.
The local search step considers only the two nearest discrete pairs around each continuous solution; if the fitted line places the true feasible pairs far from that neighborhood, a broader search (e.g., branch-and-bound over $\mathcal{F}$) would be needed to preserve performance.
The scheme assumes the jammer's power $P_J$ is known up to a bounded error; in practice an adaptive jammer that changes its beamformer or power during transmission would break the static worst-case model, motivating a fast online recalibration loop.

Share X Bluesky LinkedIn Reddit HN

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

Short version: the paper's central SDP claim doesn't survive contact with its own constraint (23e). That's the thing to know. The idea behind the paper is fine: use an IOS with coupled discrete reflection/refraction phases to null jamming and boost desired signals at the same time, and the ZF plus water-filling power allocation is standard. The novelty claim, being the first to handle discrete coupled IOS phases in anti-jamming, is fair.

But the algorithm as written is not solvable. Constraint (23e) sets each off-diagonal entry of Ξt to a nonlinear function of arg(Ξr(m1,m2)). Arg is nonconvex, the equality is nonconvex, and the feasible set isn't semidefinite-representable. Dropping the rank-one constraint doesn't make it convex. So Step 2 can't be handed to a convex solver as claimed. This is more fundamental than the reader's worry about the linear fit (21b): even with perfect s and v, the SDP story breaks. The linear fit itself is also unverified—no fitting procedure, no error measure—so the relaxation's fidelity to F is unclear. The S-procedure LMI (28)-(29) is asserted without derivation, which is a smaller but still real gap.

What's good: the problem is worth solving, the system model is carefully laid out, and the simulations, if the algorithm were valid, would show a plausible qualitative advantage over IRS baselines. The paper also honestly cites the STAR-RIS anti-jamming work [10] and explains the difference.

Who gets value: someone working on IOS or RIS anti-jamming will find the problem formulation and the decomposition useful, but not the algorithm in its current form.

Recommendation: send to peer review, with a clear expectation of major revision. The flaw is fixable: for example, alternating between Ξr and Ξt, or replacing (23e) with a proper convex surrogate. But as submitted, the central result doesn't hold.

Referee Report

3 major / 5 minor

Summary. The paper considers an intelligent omni-surface (IOS) aided downlink multi-user system that must suppress a multi-antenna jammer. The proposed scheme combines zero-forcing digital beamforming with water-filling power allocation at the base station and a three-step IOS phase-shift design: continuous relaxation with a linear coupling model, an SDP-like reformulation using a Cauchy-Schwarz bound and the S-procedure, and a local search to recover discrete coupled phases. Simulation results are reported for a 28 GHz scenario with uncertain jammer-related CSI, showing improved sum rate and reduced jamming power relative to IRS-based baselines.

Significance. If the algorithmic derivation were correct, the paper would be a relevant contribution to anti-jamming design with IOSs, since it explicitly targets the practically motivated case of discrete and coupled reflection/refraction phase shifts and includes robust jamming constraints. However, the central optimization reformulation is not a valid SDP because of the nonconvex coupling constraint (23e), and the linear coupling assumption (21b) is left unverified. These are load-bearing issues: they affect the feasibility and achievability of the reported continuous solutions and, consequently, the validity of the local search and the simulation results. The paper does not provide machine-checked proofs or reproducible code, so the numerical evidence cannot independently compensate for the mathematical gap.

major comments (3)

[III-B, Step 2, Eq. (23e)] Constraint (23e) defines the off-diagonal entries of Ξt as nonlinear functions of arg(Ξr(m1,m2)), with additional factors e^{±jv} on the last row and column. The function arg(·) is nonconvex and multivalued, and the entries of Ξr are optimization variables, so the feasible set described by (23e) is not convex and is not representable as a system of linear matrix inequalities. Consequently, problem (30), which retains (23e) together with the rank-one constraint (23c), is not an SDP, and the statement in Section III-B-ii that the problem 'can be transformed into a more tractable SDP form' is unsupported. Removing the rank-one constraint via SDR does not convexify (23e), and Algorithm 1 line 6 can therefore recover continuous phases that violate the physical coupling (21b). This flaw is load-bearing because the local search in Step 3 is initialized with infeasible continuous pairs and may restrict itself to the wrong discrete-neighbor set.
[III-B, Step 1, Eq. (21b)] The linear coupling constraint φt_m = s φr_m + v is introduced without specifying how s and v are computed from the discrete set F, what their numerical values are, or what the fitting error is. If the discrete coupled phase pairs in F do not lie on a single line in the (φr, φt) plane, the continuous relaxation either excludes feasible discrete states or includes infeasible ones. This directly affects the local search in Step 3, which selects the discrete pair for each element from only two candidates determined by the continuous solution. Since no evidence is given that the linear fit accurately represents the discrete feasible set, the relaxation step is an unverified modeling assumption that is central to the proposed method.
[III-B-iii, Eq. (29)] The conversion of the channel-uncertainty constraint (27) into the LMI (28)–(29) via the S-procedure is asserted without derivation. The paper does not show how the bounded errors ΔhJ,k and ΔHJ,k from (10)–(11) are mapped into the block matrix (29), nor that (28)–(29) are equivalent to (27) for all uncertainties in the specified bounded sets. Because this equivalence is the mechanism that makes the jamming-power constraint robust, the derivation should be supplied or a reference with a directly matching result should be cited.

minor comments (5)

[II-C] The term 'Racian model' appears in the paragraph following Eq. (4); it should be 'Rician model'.
[II-B, Eq. (3)] There is a typo in 'amplitute matrix' in the sentence following Eq. (3); it should be 'amplitude matrix'.
[IV] The simulation states 'the power ratio of the reflected and refracted signals is ε = 1', but the amplitude coefficients Γr_m and Γt_m are never defined in terms of ε; the relationship should be made explicit.
[III-B-2] Reference [14] is cited for the S-procedure, but [14] is a paper on multivariate nonnegative quadratic mappings; a standard textbook reference for the S-procedure would be more appropriate and would help readers verify the claimed conversion.
[Algorithm 1, line 6] If the SDR solution has rank greater than one and Gaussian randomization is used, the paper does not describe how many random candidates are generated, how the coupling constraint (23e) is enforced during randomization, or how the final candidate is selected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the proposed IOS-aided beamforming algorithm is a standard SDR/SCA construction evaluated by simulation, and the cited prior IOS work supplies background models rather than load-bearing conclusions.

full rationale

The paper's derivation chain is not circular. It formulates a sum-rate maximization problem (15), decomposes it into a ZF digital beamforming subproblem (16)/(19)-(20) and an IOS analog beamforming subproblem (17), relaxes the discrete coupled phase pairs to a continuous linear coupling constraint (21b), transforms the problem into an SDR form (23)-(30), and finally recovers discrete states by local search over the neighbor set F_m (Algorithm 1). None of these steps presumes the performance conclusion: the sum-rate gains and jamming-nulling behavior are obtained by running the algorithm in simulation, not by construction from the inputs. The linear fitting of s and v in (21b) is an unstated algorithmic assumption rather than a circular step, because the fitted constraint is an input to the optimization and the paper does not claim to predict a quantity that is itself fitted. The references to the authors' prior IOS work ([5], [6], [11], [12]) provide standard channel and IOS response models; they are not invoked as a uniqueness theorem or to forbid alternatives, and the central claim does not reduce to them. The performance comparisons against IRS and no-RIS baselines are conventional simulation benchmarks. Therefore no specific circular reduction can be exhibited, and the appropriate finding is no significant circularity.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The ledger shows one hand-fitted quantity (the linear coupling line) and several domain assumptions about channel error bounds and the IOS model. No new physical entities are introduced; the contribution is algorithmic. The linear coupling line is the only parameter whose value and fitting procedure are absent, and it is load-bearing.

free parameters (1)

Linear coupling coefficients s and v = not reported
Introduced in Eq. (21b) as a linear fit between reflection and refraction phase shifts across the discrete set F. The paper does not describe the fitting procedure, the fitted values, or the fit quality. Every subsequent step, including the SDP relaxation and the discrete local search, depends on this line. It is a hand-chosen parameter rather than a derived quantity.

assumptions (6)

domain assumption IOS element responses are g_r^m = Γ_r^m e^{-jφ_r^m} and g_t^m = Γ_t^m e^{-jφ_t^m} with fixed amplitudes (Eq. 2).
Taken from the authors' prior IOS circuit model [6]; no independent physical verification in this paper.
domain assumption Jamming channel estimation errors are bounded by known constants ε_d,k and ε_J,k (Eqs. 10-11).
The robust constraints rely on exact known bounds; if the actual errors exceed them, the jamming threshold is not guaranteed.
ad hoc to paper The coupled discrete phase pairs can be approximated by the global linear relation φ_t = s φ_r + v (Eq. 21b).
This is the load-bearing assumption of the paper. It is asserted as a 'linear fitting' result but no fitting method, values, or accuracy are reported.
standard math The S-procedure (cf. [14]) is applicable to the uncertainty set in (27) to yield the LMI (29).
Standard robust optimization result; the paper states (29) without derivation, assuming the uncertainty descriptor matches the required form.
domain assumption The jammer's total power estimate satisfies |P_J - P̂_J| / P_J ≤ ε_PJ, and the Cauchy-Schwarz bound (22) upper-bounds J_k.
Used to eliminate the unknown jammer beamformer v_J while keeping the constraint conservative.
standard math The SDR solution can be converted to a rank-one phase vector via Gaussian randomization [15].
Standard practice; no guarantee is given in the paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Simultaneous Beamforming and Anti-Jamming With Intelligent Omni-Surfaces." pith.science (2026). https://pith.science/paper/XH2KS4FT

@misc{pith2026250202055,
  author       = {Pith},
  title        = {Pith review of: Simultaneous Beamforming and Anti-Jamming With Intelligent Omni-Surfaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XH2KS4FT}},
  note         = {Machine review of arXiv:2502.02055}
}

read the original abstract

Wireless transmission is vulnerable to malicious jamming attacks due to the openness of wireless channels, posing a severe threat to wireless communications. Current anti-jamming studies primarily focus on either enhancing desired signals or mitigating jamming, resulting in limited performance. To address this issue, intelligent omni-surface (IOS) is a promising solution. By jointly designing its reflective and refractive properties, the IOS can simultaneously nullify jamming and enhance desired signals. In this paper, we consider an IOS-aided multi-user anti-jamming communication system, aiming to improve desired signals and nullify jamming by optimizing IOS phase shifts and transmit beamforming. However, this is challenging due to the coupled and discrete IOS reflection and refraction phase shifts, the unknown jammer's beamformer, and imperfect jammer-related channel state information. To tackle this, we relax IOS phase shifts to continuous states and optimize with a coupling-aware algorithm using the Cauchy-Schwarz inequality and S-procedure, followed by a local search to recover discrete states. Simulation results show that the proposed scheme significantly improves the sum rate amid jamming attacks.

Figures

Figures reproduced from arXiv: 2502.02055 by the authors.

**Figure 2.** (a) surface size vs. sum rate (b = 2). (b) quantization bits vs. sum rate (M = 400). (c) Cumulative distribution function of user-received jamming power with and without IOS. jammer to the users on one side for cancellation. 3)No RIS: All users receive the desired signal and the jamming signal through direct links [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

[10]

Robust beamforming design for STAR-RIS-assisted anti- jamming and secure transmission,

T. Zhou et al., “Robust beamforming design for STAR-RIS-assisted anti- jamming and secure transmission,” IEEE Trans. Green Commun. Netw. , vol. 8, no. 1, pp. 345–361, Mar. 2024

work page 2024
[1]

Hybrid beamforming for terahertz wireless communica- tions with beam squint: A survey,

Y . Wu et al. , “Hybrid beamforming for terahertz wireless communica- tions with beam squint: A survey,” Chin. J. Electron. , vol. 31, no. 6, pp. 1043–1052, Nov. 2022

work page 2022
[2]

Integrated intelligent electromagnetic radiator design for future THz communication: A review,

X. Lu et al. , “Integrated intelligent electromagnetic radiator design for future THz communication: A review,” Chin. J. Electron., vol. 31, no. 3, pp. 499–515, May. 2022

work page 2022
[3]

Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,

M. A. ElMossallamy, H. Zhang, L. Song, K. G. Seddik, Z. Han and G. Y . Li, “Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,” IEEE Trans. Cogn. Commun. Netw., vol. 6, no. 3, pp. 990–1002, Sep. 2020

work page 2020
[4]

Reconfigurable intelligent surface (RIS) assisted wireless cverage extension: RIS orientation and location optimization,

S. Zeng et al., “Reconfigurable intelligent surface (RIS) assisted wireless cverage extension: RIS orientation and location optimization,” IEEE Commun. Lett. , vol. 25, no. 1, pp. 269–273, Jan. 2021

work page 2021
[5]

Intelligent omni-surfaces for full-dimensional wireless communications: Principles, technology, and implementation,

H. Zhang et al., “Intelligent omni-surfaces for full-dimensional wireless communications: Principles, technology, and implementation,” IEEE Commun. Mag. , vol. 60, no. 2, pp. 39–45, Feb. 2022

work page 2022
[6]

Intelligent omni-surfaces: Reflection-refraction circuit model, full-dimensional beamforming, and system implementation,

S. Zeng et al. , “Intelligent omni-surfaces: Reflection-refraction circuit model, full-dimensional beamforming, and system implementation,” IEEE Trans. Commun. , vol. 70, no. 11, pp. 7711–7727, Nov. 2022

work page 2022
[7]

Intelligent reflecting surface assisted anti-jamming communications: A fast reinforcement learning approach,

H. Yang et al. , “Intelligent reflecting surface assisted anti-jamming communications: A fast reinforcement learning approach,” IEEE Trans. Wireless Commun., vol. 20, no. 3, pp. 1963–1974, Mar. 2021

work page 1963

Show all 15 references

[8]

Intelligent reflecting surface enhanced secure transmission against both jamming and eavesdropping attacks,

Y . Sun et al., “Intelligent reflecting surface enhanced secure transmission against both jamming and eavesdropping attacks,” IEEE Trans. V eh. Technol., vol. 70, no. 10, pp. 11 017–11 022, Oct. 2021

2021
[9]

Outage constrained robust beamforming optimization for multiuser IRS-assisted anti-jamming communications with incomplete information,

Y . Sun, K. An, J. Luo, Y . Zhu, G. Zheng, and S. Chatzinotas, “Outage constrained robust beamforming optimization for multiuser IRS-assisted anti-jamming communications with incomplete information,” IEEE In- ternet Things , vol. 9, no. 15, pp. 13 298–13 314, Aug. 2022

2022
[11]

Intelligent omni-surfaces: Ubiquitous wireless trans- mission by reflective-refractive metasurfaces,

S. Zhang et al. , “Intelligent omni-surfaces: Ubiquitous wireless trans- mission by reflective-refractive metasurfaces,” IEEE Trans. Wireless Commun., vol. 21, no. 1, pp. 219–233, Jan. 2022

2022
[12]

Hybrid beamforming for reconfigurable intelligent surface based multi-user communications: Achievable rates with limited discrete phase shifts,

B. Di, H. Zhang, L. Song, Y . Li, Z. Han, and H. V . Poor, “Hybrid beamforming for reconfigurable intelligent surface based multi-user communications: Achievable rates with limited discrete phase shifts,” IEEE J. Sel. Areas Commun. , vol. 38, no. 8, pp. 1809–1822, Aug. 2020

2020
[13]

Prospects and challenges of THz precoding,

R. Guo et al. , “Prospects and challenges of THz precoding,” Chin. J. Electron., vol. 31, no. 3, pp. 488–498, May. 2022

2022
[14]

Multivariate nonnegative quadratic mappings,

Z. Luo, J. Sturm, and S. Zhang, “Multivariate nonnegative quadratic mappings,” SIAM J. Optimiz. , vol. 14, no. 4, pp. 1140–1162, 2004

2004
[15]

Semidefinite relaxation of quadratic optimization prob- lems,

Z. Luo et al. , “Semidefinite relaxation of quadratic optimization prob- lems,” IEEE Signal Process. Mag. , vol. 27, no. 3, pp. 20–34, 2010

2010

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.

[1] [10]

Robust beamforming design for STAR-RIS-assisted anti- jamming and secure transmission,

T. Zhou et al., “Robust beamforming design for STAR-RIS-assisted anti- jamming and secure transmission,” IEEE Trans. Green Commun. Netw. , vol. 8, no. 1, pp. 345–361, Mar. 2024

work page 2024

[2] [1]

Hybrid beamforming for terahertz wireless communica- tions with beam squint: A survey,

Y . Wu et al. , “Hybrid beamforming for terahertz wireless communica- tions with beam squint: A survey,” Chin. J. Electron. , vol. 31, no. 6, pp. 1043–1052, Nov. 2022

work page 2022

[3] [2]

Integrated intelligent electromagnetic radiator design for future THz communication: A review,

X. Lu et al. , “Integrated intelligent electromagnetic radiator design for future THz communication: A review,” Chin. J. Electron., vol. 31, no. 3, pp. 499–515, May. 2022

work page 2022

[4] [3]

Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,

M. A. ElMossallamy, H. Zhang, L. Song, K. G. Seddik, Z. Han and G. Y . Li, “Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,” IEEE Trans. Cogn. Commun. Netw., vol. 6, no. 3, pp. 990–1002, Sep. 2020

work page 2020

[5] [4]

Reconfigurable intelligent surface (RIS) assisted wireless cverage extension: RIS orientation and location optimization,

S. Zeng et al., “Reconfigurable intelligent surface (RIS) assisted wireless cverage extension: RIS orientation and location optimization,” IEEE Commun. Lett. , vol. 25, no. 1, pp. 269–273, Jan. 2021

work page 2021

[6] [5]

Intelligent omni-surfaces for full-dimensional wireless communications: Principles, technology, and implementation,

H. Zhang et al., “Intelligent omni-surfaces for full-dimensional wireless communications: Principles, technology, and implementation,” IEEE Commun. Mag. , vol. 60, no. 2, pp. 39–45, Feb. 2022

work page 2022

[7] [6]

Intelligent omni-surfaces: Reflection-refraction circuit model, full-dimensional beamforming, and system implementation,

S. Zeng et al. , “Intelligent omni-surfaces: Reflection-refraction circuit model, full-dimensional beamforming, and system implementation,” IEEE Trans. Commun. , vol. 70, no. 11, pp. 7711–7727, Nov. 2022

work page 2022

[8] [7]

Intelligent reflecting surface assisted anti-jamming communications: A fast reinforcement learning approach,

H. Yang et al. , “Intelligent reflecting surface assisted anti-jamming communications: A fast reinforcement learning approach,” IEEE Trans. Wireless Commun., vol. 20, no. 3, pp. 1963–1974, Mar. 2021

work page 1963