Pith. sign in

REVIEW 4 major objections 6 minor 23 references

Circuit-Compliant Optimization of Dynamic Metasurface Antennas for Near-Field Localization

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Configuring a DMA receiver's analog phases with a circuit model that includes mutual coupling improves near-field localization accuracy, and closed-form phase updates keep the optimization computationally light.

desk verdict Timely and genuinely new framework for mutual-coupling-aware DMA localization, but the paper evaluates its optimized weights with the same truncated model used to derive them, so the performance claim needs an exact-model check. read the letter →

arxiv 2411.09362 v1 pith:FMUQVN7C submitted 2024-11-14 eess.SP

Ioannis Gavras , George C. Alexandropoulos This is my paper · ORCID
classification eess.SP
keywords dynamicmetasurfaceantennasnear-fieldlocalizationCramér-Raoboundmutualcouplinghybridbeamformingpositionerrorclosed-formoptimizationRayleighquotient
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 tries to show that a Dynamic Metasurface Antenna (DMA) receiver, when configured using a circuit-compliant model that accounts for mutual coupling between its densely packed metamaterial elements, can localize a nearby user more accurately than conventional multi-antenna receivers designed under idealized models. The authors derive the Cramér-Rao bound for user position estimation through the DMA's analog and digital beamforming stages, and turn the bound's minimization into a Rayleigh quotient optimization. Using a Taylor/Woodbury expansion of the circuit model's inverse matrix, they obtain first- and second-order approximations of the analog beamforming matrix, and give closed-form per-element phase solutions for both. Simulations show the second-order, coupling-aware design tracks the position error bound closely and outperforms idealized DMA and hybrid beamforming baselines, especially for dense apertures.

What carries the argument

The enabling object is the Woodbury-Taylor expansion of $(\mathbf{W}_{\mathrm{TA}} + \mathbf{W}_{\mathrm{MC}})^{-1} \approx -\mathbf{W}_{\mathrm{MC}}^{-1} \sum_{n=1}^{\infty} (-1)^n (\mathbf{W}_{\mathrm{MC}}\mathbf{W}_{\mathrm{TA}}^{-1})^n$, truncated at first or second order to yield tractable expressions for the analog beamforming matrix. The design objective is the Rayleigh quotient $\mathbf{v}^H \mathbf{W}_{\mathrm{RX}} \mathbf{A} \mathbf{W}_{\mathrm{RX}}^H \mathbf{v} / (\mathbf{v}^H \mathbf{W}_{\mathrm{RX}} \mathbf{W}_{\mathrm{RX}}^H \mathbf{v})$ with $\mathbf{A} = \sum_i (\partial \mathbf{h}/\partial \zeta_i)(\partial \mathbf{h}/\partial \zeta_i)^H$, whose unconstrained optimum is the dominant singular vector; the Lorentzian phase constraint is then enforced by per-element least-squares phase selection. For the first-order model each phase is independent, giving the closed form $\phi_n = \arg\min_{x\in\{-\pi/2, c_n, \pi/2\}} f_1(x)$; for the second-order model coupling appears through a term $b$, producing three critical points per element. These closed forms are what make coupling-aware optimization feasible for large arrays.

What would settle it

Compute the exact $(\mathbf{W}_{\mathrm{TA}} + \mathbf{W}_{\mathrm{MC}})^{-1}$ for the optimized phase configurations from (19) across the simulation settings (various $N_{\mathrm{RF}}$, $N_E$, and frequencies around 28 GHz) and compare the squared Frobenius norm difference to the second-order approximation. If this difference is not small relative to the PEB gaps shown in Figs. 2 and 3, the claimed performance advantage does not follow from the true circuit model.

Watch

Extended reading notes

Core claim

The paper's central claim is that accounting for mutual coupling in the DMA's analog beamforming optimization improves near-field localization accuracy, and that this can be done with closed-form phase updates. Under the circuit-compliant model $\mathbf{W}_{\mathrm{RX}} = \mathbf{P}_{\mathrm{SA}}^{H}(\mathbf{W}_{\mathrm{TA}} + \mathbf{W}_{\mathrm{MC}})^{-1}$, the first-order approximation $\mathbf{W}_{\mathrm{RX}} \approx \mathbf{P}_{\mathrm{SA}}^{H}\mathbf{W}_{\mathrm{TA}}^{-1}$ ignores coupling entirely, while the second-order approximation keeps a coupling term: $\mathbf{W}_{\mathrm{RX}} \approx \mathbf{P}_{\mathrm{SA}}^{H}(\mathbf{W}_{\mathrm{TA}}^{-1} - \mathbf{W}_{\mathrm{TA}}^{-1}\mathbf{W}_{\mathrm{MC}}\mathbf{W}_{\mathrm{TA}}^{-1})$. The authors show that minimizing a tight lower bound on the position error bound is equivalent to maximizing a Rayleigh quotient, whose optimal rank-one solution can be decomposed into digital weights $\mathbf{v}$ and per-element analog phases via least squares. The resulting phase solutions, given in equations (13) and (19), are simple trigonometric critical points, making the approach computationally light even for extremely large apertures. Numerical results indicate that, as the number of RF chains grows, the second-order DMA design approaches the idealized-model performance while remaining robust to coupling.

Load-bearing premise

The Taylor and Woodbury expansion of $(\mathbf{W}_{\mathrm{TA}} + \mathbf{W}_{\mathrm{MC}})^{-1}$, truncated at second order, is assumed to be accurate for all relevant array sizes, frequencies, and phase configurations; if the truncation error is large in the optimized operating regime, the optimized weights do not correspond to the true circuit model.

Editorial extensions

If this is right

  • A DMA receiver can be configured for near-field localization without pretending mutual coupling does not exist, using only per-element phase computations.
  • The second-order, coupling-aware design should be preferred when the number of RF chains is small, while the first-order design becomes a low-cost alternative as the aperture grows.
  • The derived PEB expression and the Rayleigh quotient reformulation give a ready-made objective for other DMA-based estimation tasks, such as angle-of-arrival or range estimation.
  • The gap between idealized and coupling-aware models shrinks as the number of RF chains increases, suggesting that idealized models are most misleading for small-NRF dense arrays.

Reading between the lines

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

  • The same per-element phase-splitting trick may extend to higher-order approximations or to other hardware constraints, since the least-squares decomposition treats the digital vector as a free parameter.
  • If the approximation-error behavior shown in Fig. 1 holds in the optimized regime, the first-order model may be sufficient for extremely large apertures, making the coupling term a controlled correction rather than a necessity.
  • A natural next test is multi-user localization or joint sensing and communication with the same coupling-aware objective, where the CRB minimization would need to compete against multiuser interference.
  • The framework assumes known channel derivatives; in practice these would be estimated from received pilots, so the robustness of the closed-form phases to channel estimation error is a testable extension.
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.

Referee Report

4 major / 6 minor

Summary. This manuscript considers a DMA receiver for near-field localization of a single-antenna UE. The authors adopt the circuit-compliant model W_RX = P_SA^H (W_TA + W_MC)^-1 from Williams et al., expand the inverse using a Woodbury/Taylor series, and retain first- and second-order terms. They derive the FIM/CRB/PEB for the user position, formulate the minimization of a lower bound on the PEB as a Rayleigh quotient maximization, and project the dominant singular vector onto the Lorentzian-constrained analog weights, yielding per-element phase rules (13) and (19). The paper claims that the proposed mutual-coupling-aware DMA design outperforms idealized DMA and hybrid A/D baselines in RMSE and PEB.

Significance. If the central claim is supported, the paper would be a useful contribution: it is the first DMA localization design, to my knowledge, that incorporates mutual coupling through a physically motivated circuit model rather than a fitted surrogate; the objective is derived from the CRB, not from data fitting; the optimization respects the Lorentzian weight constraint; and the paper clearly distinguishes first- and second-order approximation costs. The main strengths are the external electromagnetic model import from [18], the explicit CRB-based objective, and the closed-form per-element phase updates. However, the validation does not yet demonstrate that the optimized weights are near-optimal for the exact circuit-compliant model, so the significance is conditional on the model-approximation and projection issues detailed below.

major comments (4)
  1. [Section II-A and Section IV] The Taylor expansion of (I_N + W_MC W_TA^{-1})^{-1} is used without stating or verifying its convergence condition. The Neumann series converges only when the spectral radius of W_MC W_TA^{-1} is less than one, and the paper neither proves this for the physical model nor checks it for the optimized phase profiles. Figure 1 validates the inverse approximation only for NRF=4, NE=32 and unspecified phase configurations, while the simulations in Figs. 2 and 3 use NRF=2 with NE=128 or 256 and the optimized phase profiles. More importantly, the RMSE and PEB results appear to be computed with the same truncated W_RX used in the optimization; I found no evaluation of the optimized weights against the exact inverse P_SA^H (W_TA + W_MC)^{-1}. Consequently, the central claim that the proposed design maintains accuracy under mutual coupling is not yet established. I recommend adding a spectral-radius check or an error bound for the optimized configurations, and a Monte Carlo/CRB comparison in which the optimized weights are evaluated with the exact circuit-compliant model.
  2. [Section II (WMC definition) and Section III-B2] The second-order solution uses diagonal entries of W_MC, but these entries are not defined by the manuscript. The matrix W_MC is specified through the Green's functions G_MC and G_SA, and G_MC contains terms involving 1/R and 1/R^2 with R = ||p_n - p_{n'}||. For n = n', R = 0, so the diagonal entry [W_MC]_{n,n} is either undefined or singular, and the same issue affects G_SA(p_n, p_n). Yet Eq. (16) uses [W_MC^H]_{n,n} and Eq. (17) uses [WMC]_{n,n}. The paper needs to state how the self-coupling terms are regularized or excluded; without this, the second-order phase solution is not computable as written.
  3. [Section III-B (OP and OP1)] The transformation from OP to OP1 is not equivalence-preserving. OP maximizes the Rayleigh quotient over e = W_RX^H v, whereas OP1 minimizes the Euclidean distance to a fixed dominant singular vector e_opt = u1 sqrt(sigma1). Because the Rayleigh quotient is scale-invariant but the least-squares objective is not, the arbitrary scaling sqrt(sigma1) changes the solution, and the projection need not preserve a high quotient value. The paper does not report the achieved quotient (or the resulting FIM/PEB) of the projected weights relative to the unconstrained optimum sigma1, nor does it compare against a direct optimization of the original objective. This gap is load-bearing for the claim that the closed-form phase rules are near-optimal; I recommend adding such an assessment or describing OP1 explicitly as a heuristic.
  4. [Section III-B2, Eqs. (16)-(19)] The derivation of the critical points is omitted (the paper refers to 'standard numerical tools') and the expression in (17) is presented without justification. Since f2 depends on b, which couples all elements, the per-element variables are not decoupled; the paper itself states that optimality cannot be guaranteed. The iterative procedure is described only as 'until a stopping criterion is met', with no convergence guarantee, initialization strategy, or number of iterations reported. Given that the second-order solution is one of the paper's two main contributions, I ask for the derivation or a citation for (17), a specification of the solver and stopping rule, and empirical convergence behavior for the simulated configurations.
minor comments (6)
  1. [Section IV, Fig. 2] The Fig. 2 caption states NE = 256, but the text in Section IV says NE = 128; please reconcile this discrepancy.
  2. [Section IV] In the paragraph discussing Fig. 3, 'A shown' should read 'As shown'.
  3. [Section II-C and Section III-A] The dimensions in the signal model are unclear: h is defined as an N-dimensional column vector, but Eq. (6) writes h^H s and the derivatives in (8) and (9) mix h and h^H. Please clarify the orientation of h, s, and the received signal so that the FIM derivation is unambiguous.
  4. [Section II (channel model)] The radiation profile F(theta_i,n) is only referenced to [13, eq. (6)] and not reproduced; without an explicit expression the numerical results are not fully reproducible.
  5. [Section III-B] The bound Tr{I^{-1}} >= 9T^2/Tr{I} is not universally tight, so OP maximizes a surrogate rather than the exact PEB. Please clarify this and report the gap between the surrogate and the true PEB for the simulated designs.
  6. [Section II, Eq. (2)] The codebook W is called a 'phase profile codebook', but |0.5(j + e^{j phi})| varies with phi; please adjust the wording to make clear that amplitude and phase are jointly tuned.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the optimization objective and closed-form analog weights are derived analytically from an externally imported circuit model, with no fitted parameter renamed as a prediction.

full rationale

The paper's derivation chain is self-contained in the sense required by the circularity check. The circuit-compliant DMA model WRX = P_SA^H (WTA + WMC)^{-1} is imported from an external prior work [18], not constructed inside the paper to force the later results. The Woodbury/Taylor expansion of (WTA + WMC)^{-1} is a standard mathematical approximation; its accuracy is checked in Fig. 1 against the exact inverse for a sample configuration, and no parameter is fitted to the PEB or RMSE curves. The CRB and PEB in (9)-(10) are derived analytically from the assumed Gaussian signal model, and the objective OP is a Rayleigh quotient maximization whose unconstrained optimum is the principal singular vector of A. The closed-form phase solutions (13) and (19) minimize the least-squares distance to that unconstrained optimum under the Lorentzian codebook; they are not equal by construction to the plotted PEB or RMSE. The numerical evaluation does use the same truncated WRX in both the optimized weights and the CRB, which is a legitimate concern about whether the truncated model matches the exact circuit model in the simulated operating regime, but that is a model-validation or correctness risk, not circularity: the prediction is not manufactured from a fitted parameter or from a self-citation. The self-citations ([13], [14]) provide the Lorentzian constraint form and baseline architectures, not the load-bearing uniqueness or correctness of the mutual-coupling model, and the key physical model [18] is external. Accordingly, no circular step is identified.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no fitted free parameters and no new physical entities. It imports the circuit model, channel model, and Lorentzian codebook from prior work, and adds a Taylor approximation whose convergence is only numerically spot-checked. The main risk is not circularity but whether the approximated circuit model is accurate in the optimized regime.

assumptions (5)
  • domain assumption The circuit-compliant DMA model W_RX = P_SA^H (W_TA + W_MC)^{-1} from [18] accurately describes the physical DMA reception behavior.
    This is the foundational model imported from [18]; the paper does not re-derive or experimentally validate it. It enters in Section II, Eq. (1).
  • ad hoc to paper The Taylor expansion of (W_TA + W_MC)^{-1} converges and the first- or second-order truncation remains accurate.
    No convergence condition is proved; accuracy is only checked numerically in Fig. 1 for a few configurations (NRF = 4, NE = 32, selected frequencies). This assumption is introduced to make optimization tractable.
  • domain assumption The near-field channel model h in Eq. (3) with radiation profile F from [13, eq. (6)] fully characterizes UE-to-element propagation, with mutual coupling handled only in W_RX.
    The channel vector uses a standard spherical-wave model and an external radiation profile; the paper does not justify the separation of coupling effects between channel and circuit model. Section II-B, Eq. (3).
  • ad hoc to paper The unconstrained Rayleigh-quotient solution ev_opt remains nearly achievable after least-squares projection onto the Lorentzian-constrained DMA structure.
    The original constrained problem is relaxed by dropping the Lorentzian constraint, solving OP, and then projecting onto the feasible set via OP1; no optimality gap is quantified. Section III-B.
  • standard math The complex Gaussian FIM expression and the inequality Tr{I^{-1}} >= 9T^2/Tr{I} are applicable to this estimation problem.
    Standard estimation theory [21] is used; the inequality converts PEB minimization into Rayleigh quotient maximization. Section III-A.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Circuit-Compliant Optimization of Dynamic Metasurface Antennas for Near-Field Localization." pith.science (2026). https://pith.science/paper/FMUQVN7C

@misc{pith2026241109362,
  author       = {Pith},
  title        = {Pith review of: Circuit-Compliant Optimization of Dynamic Metasurface Antennas for Near-Field Localization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FMUQVN7C}},
  note         = {Machine review of arXiv:2411.09362}
}
read the original abstract

This paper presents an optimization framework for near-field localization with Dynamic Metasurface Antenna (DMA) receivers. This metasurface technology offers enhanced angular and range resolution realizing efficient hybrid Analog and Digital (A/D) BeamForming (BF) with sub-wavelength-spaced metamaterials of tunable responses. However, the vast majority of the state-of-the-art DMA designs is based on an idealized model for their reception operation, which neglects several practical aspects, such as the inevitable mutual coupling among the densely deployed metamaterials within a given aperture. Capitalizing on a recent circuit-compliant active metasurface model, we present a novel mutual-coupling-aware framework for localization-optimized hybrid A/D BF weights at the reception DMA. To deal with the intrinsic complexity of the deployed model, we introduce first- and second-order approximations for the DMA analog BF matrix that enable efficient optimization, while maintaining accuracy. We derive the Cramer-Rao Bound for the user position estimation which serves as our design objective for the hybrid A/D BF matrices. Closed-form solutions for these matrices for both approximations are presented, whose validity is confirmed via numerical investigations. It is also demonstrated that the proposed DMA design outperforms state-of-the-art multi-antenna reception architectures optimized for the same localization objective.

Figures

Figures reproduced from arXiv: 2411.09362 by the authors.

Figure 1
Figure 1. Squared norm difference between the approximated and actual values of (WTA+WMC) −1 as a function of the approximation order for various operating frequencies f, considering an RX DMA panel with NRF = 4 microstrips each with NE = 32 metamaterials. Next, we apply the Taylor series expansion to the term WMCW−1 TA + IN −1 , yielding: WMCW−1 TA + IN −1 ≈ IN + X∞ n=1 (−1)n WMCW−1 TAn . Combining the previously derived … view at source ↗
Figure 2
Figure 2. RMSE of the UE’s position versus Pmax, considering an RX DMA equipped with NRF = 2 microstrips, each hosting NE = 256 phase-tunable metamaterials. chosen as λ/5, whereas, for the latter, as λ/2. Notably, none of those model accounts for mutual coupling between adjacent antennas/metamaterials. In [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. PEB of the UE’s position versus the number of RX RF chains, considering an RX DMA with NE = 256 phase-tunable metamaterials per microstrip and Pmax = −12 dBm. demonstrate better scalability in terms of PEB performance compared to the HBF architecture, and outperform the HBF structure, despite the presence of mutual coupling. A shown, when the number of RX RF chains increases, the performance gap between the second-o… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 19 canonical work pages

  1. [18]

    Electromagnetic based communication model for dynamic metasurface antennas,

    R. J. Williams et al., “Electromagnetic based communication model for dynamic metasurface antennas,” IEEE Trans. Wireless Commun., vol. 21, no. 10, pp. 8616–8630, 2022

    work page 2022

  2. [1]

    A tutorial on extremely large-scale MIMO for 6G: Fundamentals, signal processing, and applications,

    Z. Wang et al. , “A tutorial on extremely large-scale MIMO for 6G: Fundamentals, signal processing, and applications,” IEEE Commun. Surveys & Tuts. , early access, 2024

    2024

  3. [2]

    Holographic MIMO communications: Theoretical foun- dations, enabling technologies, and future directions,

    T. Gong et al., “Holographic MIMO communications: Theoretical foun- dations, enabling technologies, and future directions,” IEEE Commun. Surveys & Tuts. , vol. 26, no. 1, pp. 196–257, 2024

    2024

  4. [3]

    Reconfigurable intelligent surfaces for 6G: Emerging hardware architectures, applications, and open challenges,

    E. Basar et al. , “Reconfigurable intelligent surfaces for 6G: Emerging hardware architectures, applications, and open challenges,” IEEE V eh. Technol. Mag., vol. 19, no. 3, pp. 27–47, 2024

    work page 2024

  5. [4]

    Hybrid reconfigurable intelligent meta- surfaces: Enabling simultaneous tunable reflections and sensing for 6G wireless communications,

    G. C. Alexandropoulos et al. , “Hybrid reconfigurable intelligent meta- surfaces: Enabling simultaneous tunable reflections and sensing for 6G wireless communications,” IEEE V eh. Technol. Mag., vol. 19, no. 1, pp. 75–84, 2024

    work page 2024

  6. [5]

    Dynamic metasurface antennas for 6G extreme massive MIMO communications,

    N. Shlezinger et al. , “Dynamic metasurface antennas for 6G extreme massive MIMO communications,” IEEE Wireless Comm., vol. 28, no. 2, pp. 106–113, 2021

    work page 2021

  7. [6]

    Energy efficiency maximization of massive MIMO communications with dynamic metasurface antennas,

    L. You et al. , “Energy efficiency maximization of massive MIMO communications with dynamic metasurface antennas,” IEEE Trans. Wireless Commun., vol. 22, no. 1, pp. 393–407, 2023

    work page 2023

  8. [7]

    Metasurface-based receivers with 1-bit ADCs for multi-user uplink communications,

    P. Gavriilidis et al. , “Metasurface-based receivers with 1-bit ADCs for multi-user uplink communications,” in Proc. IEEE ICASSP, Seoul, South Korea, 2024, pp. 9141–9145

    work page 2024

Show all 23 references
  1. [8]

    Near-field beam tracking with extremely massive dynamic metasurface antennas,

    P. Gavriilidis and G. C. Alexandropoulos, “Near-field beam tracking with extremely massive dynamic metasurface antennas,” arXiv preprint arXiv:2406.01488, 2024

    2024 arXiv

  2. [9]

    Beam focusing for near-field multiuser mimo com- munications,

    H. Zhang et al. , “Beam focusing for near-field multiuser mimo com- munications,” IEEE Trans. Wireless Commun. , vol. 21, no. 9, pp. 7476– 7490, 2022

    2022

  3. [10]

    Near-field localization with 1-bit quantized hybrid A/D reception,

    I. Gavras et al. , “Near-field localization with 1-bit quantized hybrid A/D reception,” in Proc. IEEE ICASSP , Seoul, South Korea, 2024, pp. 13 286–13 290

    2024

  4. [11]

    Autoregressive attention neural networks for non-line-of-sight user tracking with dynamic metasurface antennas,

    K. Stylianopoulos et al. , “Autoregressive attention neural networks for non-line-of-sight user tracking with dynamic metasurface antennas,” in Proc. IEEE CAMSAP , Los Sue ˜nos, Costa Rica, 2023

    2023

  5. [12]

    Near-field localization with dynamic metasurface antennas,

    Q. Yang et al. , “Near-field localization with dynamic metasurface antennas,” in Proc. IEEE ICASSP , Rhodes Island, Greece, 2023

    2023

  6. [13]

    Full duplex holographic MIMO for near-field integrated sensing and communications,

    I. Gavras et al., “Full duplex holographic MIMO for near-field integrated sensing and communications,” in Proc. EUSIPCO , Helsinki, Finland, 2023

    2023

  7. [14]

    Simultaneous near-field THz communications and sensing with full duplex metasurface transceivers,

    I. Gavras and G. C. Alexandropoulos, “Simultaneous near-field THz communications and sensing with full duplex metasurface transceivers,” in Proc. IEEE SPA WC, Lucca, Italy, 2024

    2024

  8. [15]

    Joint near-field target tracking and communications with full duplex holographic MIMO,

    ——, “Joint near-field target tracking and communications with full duplex holographic MIMO,” in Proc. IEEE ICASSP, Seoul, South Korea, 2024

    2024

  9. [16]

    Integrated sensing and communication with mil- limeter wave full duplex hybrid beamforming,

    M. A. Islam et al. , “Integrated sensing and communication with mil- limeter wave full duplex hybrid beamforming,” in Proc. IEEE Int. Conf. Commun., Seoul, South Korea, 2022, pp. 4673–4678

    2022

  10. [17]

    Physically consistent models for intelligent reflective surface-assisted communications under mutual coupling and element size constraint,

    M. Akrout et al., “Physically consistent models for intelligent reflective surface-assisted communications under mutual coupling and element size constraint,” in Proc. Asilomar Conf. Signals, Sys., Comp. , 2023, pp. 1589–1594

    2023

  11. [19]

    Physfad: Physics-based end-to-end channel modeling of RIS-parametrized environments with adjustable fading,

    R. Faqiri et al. , “Physfad: Physics-based end-to-end channel modeling of RIS-parametrized environments with adjustable fading,” IEEE Trans. Wireless Commun., vol. 22, no. 1, pp. 580–595, 2023

    2023

  12. [20]

    On the tacit linearity assumption in common cascaded models of RIS-parametrized wireless channels,

    A. Rabault et al., “On the tacit linearity assumption in common cascaded models of RIS-parametrized wireless channels,” IEEE Trans. Wireless Commun., vol. 23, no. 8, pp. 10 001–10 014, 2024

    2024

  13. [21]

    S. M. Kay, Fundamentals of statistical signal processing: estimation theory. Prentice-Hall, Inc., 1993

    1993

  14. [22]

    Tutorial on maximum likelihood estimation,

    I. J. Myung, “Tutorial on maximum likelihood estimation,” J. Math. Psychology, vol. 47, no. 1, 2003

    2003

  15. [23]

    Low-complexity and high-resolution DOA estimation for hybrid analog and digital massive MIMO receive array,

    F. Shu et al., “Low-complexity and high-resolution DOA estimation for hybrid analog and digital massive MIMO receive array,” IEEE Trans. Commun., vol. 66, no. 6, pp. 2487–2501, 2018

    2018

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.