Pith. sign in

REVIEW 2 major objections 5 minor 36 references

Reciprocity Calibration of Dual-Antenna Repeaters via MMSE Estimation

T0 review · 2 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read This paper claims that reciprocity calibration of a dual-antenna repeater—estimating the ratio γ = β/α between the repeater's forward and reverse gains—can be recast as a Bayesian MMSE problem whose estimator, built on von Mises denoisers a

desk verdict Credible Bayesian reformulation of repeater calibration, but the γ-prior MoM step is circular as written—Algorithm 3 is not executable without a fix. read the letter →

arxiv 2602.05724 v1 pith:AJDNL2UP submitted 2026-02-05 eess.SP cs.ITmath.IT

Shoma Hara , Takumi Takahashi , Hiroki Iimori , Hideki Ochiai , Erik G. Larsson This is my paper · ORCID
classification eess.SPcs.ITmath.IT
keywords reciprocitycalibrationdual-antennarepeaterrepeater-assistedMIMOMMSEestimationvonMisesdenoisermethodofmomentsBayesianbilinearinferenceTDDchannel
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

The paper is trying to establish that reciprocity calibration of a dual-antenna repeater can be solved as Bayesian bilinear inference rather than as a deterministic least-squares problem. The proposed MMSE algorithm incorporates statistical priors for the hardware reciprocity coefficients and noise, updates parameters with von Mises denoisers, and estimates the unknown prior statistic of |γ|² via the method of moments. Simulation results show RMSE gains of roughly 4 dB for small arrays and 14 dB for large arrays over the basic NLS baseline at RMSE = 10⁻¹, with convergence in about 4 iterations. If correct, this offers a practical, low-complexity way to maintain TDD channel reciprocity in repeater-assisted massive MIMO systems.

What carries the argument

The load-bearing component is the von Mises denoiser—a closed-form Bayes-optimal estimator for a point on a circle observed in AWGN. It converts each scalar observation into a posterior mean whose magnitude is the modified-Bessel ratio I₁(·)/I₀(·) and whose phase follows the observation's phase. The paper applies this denoiser to the diagonal entries of A and B and to γ, and couples it with a method-of-moments formula that estimates the unknown statistic φ_γ = E|γ|² from the same measurement, avoiding the need for long-term calibration statistics. The overall loop uses probabilistic data association with Gaussian approximation of the residual interference terms to propagate means and MSEs.

What would settle it

Run Algorithm 3 with φ_γ set to the true value versus the MoM estimate and compare the RMSE of γ̂; alternatively, monitor the quantity (|q|² − u)/(u² + s) from equation (48) across many trials. If negative values of |γ̂|² occur at a non-negligible rate, or if the two RMSE curves diverge, then the claimed MMSE gains do not follow from the algorithm as written.

Watch

Extended reading notes

Core claim

The central claim is that the gain ratio γ = β/α can be estimated in the MMSE sense from the four effective measurements R1–R4 obtained by two bidirectional pilot transmissions with the repeater in its nominal and π-phase-shifted states. The paper's key argument is that the diagonal reciprocity matrices A and B have a phase-dominant structure captured by a von Mises prior, so their MMSE updates reduce to the closed-form denoiser η(y; v) = r·I₁(|ζ|)/I₀(|ζ|)·e^{j arg ζ}; the same denoiser is applied to γ after estimating |γ| by the method of moments. Under AWGN and standard Rayleigh-fading simulation settings, the paper reports consistently lower RMSE than NLS and, for large arrays, an error t

Load-bearing premise

The load-bearing premise is that the method-of-moments estimator in Lemma 2 can compute |γ|² without knowing |γ|², even though its formula for s in (49c) contains φ_γ = E|γ|²; if that circularity cannot be resolved by an initialization or fixed-point iteration, the final γ-MMSE output of Algorithm 3 is undefined.

Editorial extensions

If this is right

  • A dual-antenna repeater can be made effectively reciprocal by configuring its gain to compensate the estimated ratio γ̂, preserving downlink precoding in TDD repeater-assisted massive MIMO.
  • Calibration stays accurate in the low-SNR and large-array regimes where deterministic NLS saturates, because the prior acts as a regularizer on the estimation problem.
  • The algorithm converges in roughly 4 iterations, making it usable within a channel coherence interval with modest training overhead.
  • When only diagonal noise variances are used, the computational complexity is on the same order as basic NLS, so the accuracy gain does not require heavy matrix inversions.
  • The framework can exploit known colored noise through a separable Kronecker covariance structure at additional computational cost.

Reading between the lines

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

  • The phase-only prior assumption (β = 0, unit radius for A and B) is the main regularizer; a natural robustness check is to test the denoiser under non-uniform phase-error distributions or amplitude mismatches that violate approximation (26).
  • The self-referential MoM step in Lemma 2 suggests a testable fix: initialize φ_γ, iterate the estimate of |γ|², and clip negative values; comparing this fixed-point version with the paper's single-shot version would quantify how much of the reported gain depends on the unresolved initialization.
  • The same bilinear-inference-plus-circular-denoiser structure could extend to other cascaded hardware whose forward and reverse responses differ, such as RIS elements or relays, where only the ratio matters for reciprocity.
  • A direct experimental validation would be to run Algorithm 3 on measured repeater hardware with known α and β and check whether the achieved downlink beamforming gain follows the predicted 1/SNR scaling.
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

2 major / 5 minor

Summary. The paper proposes a Bayesian MMSE calibration method for dual-antenna repeaters in TDD MIMO systems. From four measurements (8), the repeater gain ratio γ=β/α and the antenna reciprocity matrices A, B are estimated. The proposed Algorithm 3 uses PDA-style alternating MMSE updates: point estimates for H and Z, von Mises denoisers for the diagonal entries of A and B with a uniform phase prior (β=0), and a MoM-based MMSE update for γ. Simulations claim 4–14 dB RMSE gains over NLS baselines at comparable complexity and convergence within about 4 iterations.

Significance. If correct, the paper would provide a practical Bayesian alternative to deterministic NLS calibration for repeater-assisted MIMO, with detailed derivations for the A/B updates (Eqs. 18–39), a self-contained von Mises denoiser derivation, and a complexity analysis distinguishing full covariance and diagonal-only implementations. However, the γ-estimation step is not executable as printed: the MoM estimator of |γ|² is self-referential, so the central output of Algorithm 3 is undefined. The claimed performance gains therefore are not supported by the printed algorithm until that step is repaired.

major comments (2)
  1. [§IV.C.4 (Lemma 2, Eqs. (48)–(49c); Algorithm 3, lines 21–24)] The MoM estimator of |γ|² is self-referential. Equation (49c) defines s = d̂ᴴΣ_A⁻¹ V_D^γ Σ_A⁻¹ d̂, and V_D^γ in (44) contains φ_γ = E|γ|². The text proposes to replace φ_γ with the MoM estimate |γ̂|², i.e. the very quantity being defined. Algorithm 3 computes V_D^γ (line 21) before any |γ̂|² exists (line 24) and gives no initialization or fixed-point update. If γ is treated as deterministic so φ_γ = |γ|², (48) becomes an implicit quadratic equation in x = |γ|²; the paper neither states nor solves this equation. Consequently Algorithm 3 is not executable, and the RMSE curves in Figs. 4 and 5 cannot be reproduced from the printed pseudocode without supplying an extra unspecified φ_γ. Please specify a well-defined procedure (e.g. initialize φ_γ and iterate the MoM update, or solve the implicit equation) and clip non-positive estimates.
  2. [§IV.C.4, Lemma 2 proof (Eqs. (50)–(52))] The estimator is called “consistent”, but the proof replaces the ensemble moment E|q|² by the single-sample quantity |q|². No averaging over repeated observations or concentration argument over the M_A M_B dimensions is provided. As written, the proof establishes only a moment-matching equation for one realization. Please state the asymptotic regime in which consistency is claimed and justify the approximation, or revise the wording.
minor comments (5)
  1. [§IV.C.4, Eq. (42)] The notation p_{vec(R4)|γ} denotes the conditional PDF of R4 given γ, but the surrounding text says “given R4”; please correct this notational inconsistency.
  2. [§IV.A.2, after Eq. (16)] The symbols \bar{Σ}_A and \bar{Σ}_B are introduced without prior definition, while Section IV later uses Σ_A = Ψ_A ⊗ Ω_A. Please define these consistently.
  3. [Algorithm 2, line 12] The γ update writes tr{(Â Ẑ ˆB)^H R4}; Eq. (14) and Algorithm 1 use tr{(Â Ẑ^T ˆB)^H R4}. The missing transpose appears to be a typo.
  4. [Eq. (53b)] Please check the placement of the square in the posterior MSE formula. The standard expression for the von Mises denoiser at radius |γ̂| is |γ̂|²(1 − (I1/I0)²), and the typesetting in (53b) is ambiguous.
  5. [Algorithm 3, line 27] Line 27 uses |γ̂| in the definition of ζ_γ, but the surrounding derivation and Eq. (54) use |γ̌|; please correct the symbol.

Circularity Check

1 steps flagged · score 8.0 of 10

The MoM estimator of |γ|² is self-referential: Eq. (48) uses s from Eq. (49c), and Eq. (44) makes s depend on the unknown φγ=E|γ|² that the MoM step is supposed to produce.

  1. self definitional [Section IV-C4, Eqs. (43)–(49) and Algorithm 3 lines 21–24]
    "Since ϕγ used in (44) is generally unavailable, its treatment requires further discussion. ... we approximate the true long-term statistic ϕγ by an instantaneous estimate of |γ|², obtained via the MoM described below. ... |ˇγ|² = (|q|²−u)/(u²+s), ... s≜ d̂HΣ_A^{-1} V_D^γ Σ_A^{-1} d̂ ... v_{γ,ij}≜ϕγ|Ẑ(j,i)|²( ˆvA,i| ˆB(j,j)|² + | ˆA(i,i)|²ˆvB,j + ˆvA,iˆvB,j )"

    Eq. (48) defines x=|γhat|² as (|q|²−u)/(u²+s), but s in Eq. (49c) is computed from V_D^γ, and V_D^γ in Eqs. (43)–(44) is proportional to the unknown φγ=E|γ|². The paper explicitly says to replace φγ by the MoM estimate of |γ|², i.e. by x itself. Thus Eq. (48) is an implicit equation x=(|q|²−u)/(u²+c·x) for some c≥0; the right-hand side depends on the quantity being estimated. Algorithm 3 lines 21–24 compute V_D^γ and s before any |γhat|² exists, with no initialization, fixed-point iteration, or clipping of negative values. Consequently the γ-MMSE update in Eq. (53), which is the main output of the algorithm, is undefined as written, and the RMSE curves in Fig. 4 depend on an unspecified extra input.

full rationale

The A/B estimation path is not circular: it uses a non-informative uniform phase prior plus the unit-modulus constraint, which is domain knowledge rather than a fitted target. The comparison with the NLS baseline [27] is a normal external benchmark, even though one coauthor is shared; the NLS model is stated independently and is not used to justify the MMSE derivation. However, the γ-estimation path is materially circular. Lemma 2's MoM estimator of |γ|² is defined through s in Eq. (49c), which depends on V_D^γ in Eq. (44), and V_D^γ contains the unknown φγ=E|γ|² that the MoM step claims to supply. The text explicitly substitutes the MoM estimate for φγ, making Eq. (48) an implicit self-referential equation. Algorithm 3 never initializes φγ before line 21, so the algorithm is not executable as printed. Since the paper's headline gains (roughly 4–14 dB in Fig. 4) are attributed to the complete γ-MMSE update, this is a load-bearing gap rather than a minor implementation detail.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The main 'free' inputs are the assumed noise covariances and the estimated second moment φ_γ; the latter is particularly fragile because the MoM procedure that estimates it also depends on it. The unit-circle constraint for A/B is a domain assumption that is exactly matched by the simulation setup. No new physical entities are posited.

free parameters (2)
  • φ_γ = E|γ|² (prior second moment of the repeater gain ratio) = |γ̂|²_MoM via Eq. (48) — self-referential
    Used in the γ update to build V_D^γ (Eq. 44) and hence s (Eq. 49c). The paper approximates it by an instantaneous MoM estimate from the same R4 observation, which makes the estimator depend on the unknown it estimates.
  • von Mises concentration parameter β = 0 (non-informative)
    Set to zero because the phase-error variance is difficult to estimate; the denoiser then reduces to a unit-modulus shrinkage ('circular-uniform denoiser') rather than a full Bayesian phase prior.
assumptions (6)
  • domain assumption The π-phase-shift configuration exactly negates the repeater path and leaves all other parameters unchanged, yielding the four preprocessed observations R1–R4 of Eq. (8).
    Inherited from [27]; invalid if the phase shift alters gains or couples with the direct path.
  • ad hoc to paper A(i,i) and B(j,j) are exactly on the unit circle (amplitude error terms in Eq. (26) are dropped; σ_ε ≪ 1), so r=1 in Lemma 1.
    The prior model (15) includes amplitude errors, but the denoiser ignores them; simulations use exactly unit-magnitude coefficients, so the assumption is not stress-tested.
  • domain assumption Estimation-error terms in the bilinear observations are jointly complex Gaussian (VGA/CLT) with the stated covariance (Eqs. 20–22, 32–34, 41–44).
    Required for closed-form MMSE updates; quality degrades for small arrays and early iterations.
  • domain assumption Observation noise is a matrix-valued complex Gaussian with separable Kronecker covariances (Eq. 16) that are known or estimable; at least the diagonal variances are available.
    Used to form V_A,i, V_B,j, and V_γ; the paper claims only diagonal elements are needed for the low-complexity variant.
  • ad hoc to paper The posterior-MSE identity v·∂η/∂y = E[|x−η(y)|²|y] (Eq. 69) remains valid when the prior hyperparameters are replaced by MoM estimates.
    Used in Eqs. (29), (38b), (53b) to report per-parameter MSE; strictly it holds only for the true MMSE estimator with a known prior, so the reported posterior MSEs are approximate.
  • domain assumption The rank-one approximation Ẑ = S{R2} is accurate enough that the γ update can treat Ẑ ≈ Z (stated at Eq. 40).
    Inherited from the NLS approach; no error propagation from Z into γ is modeled.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Reciprocity Calibration of Dual-Antenna Repeaters via MMSE Estimation." pith.science (2026). https://pith.science/paper/AJDNL2UP

@misc{pith2026260205724,
  author       = {Pith},
  title        = {Pith review of: Reciprocity Calibration of Dual-Antenna Repeaters via MMSE Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AJDNL2UP}},
  note         = {Machine review of arXiv:2602.05724}
}
read the original abstract

This paper proposes a novel Bayesian reciprocity calibration method that consistently ensures uplink and downlink channel reciprocity in repeater-assisted multiple-input multiple-output (MIMO) systems. The proposed algorithm is formulated under the minimum mean-square error (MMSE) criterion. Its Bayesian framework incorporates complete statistical knowledge of the signal model, noise, and prior distributions, enabling a coherent design that achieves both low computational complexity and high calibration accuracy. To further enhance phase alignment accuracy, which is critical for calibration tasks, we develop a von Mises denoiser that exploits the fact that the target parameters lie on the circle in the complex plane. Simulation results demonstrate that the proposed MMSE algorithm achieves substantially improved estimation accuracy compared with conventional deterministic non-linear least-squares (NLS) methods, while maintaining comparable computational complexity. Furthermore, the proposed method exhibits remarkably fast convergence, making it well suited for practical implementation.

Figures

Figures reproduced from arXiv: 2602.05724 by the authors.

Figure 1
Figure 1. Two antenna arrays, A and B, and a repeater (R). G denotes the propagation channel from A to B when R is turned off. Radio channels are represented by solid lines, whereas repeater gains are represented by dashed lines. The figure is adapted from [27]. a1, . . . , aM on its main diagonal. The conjugate, transpose, and Hermitian transpose are denoted by (·) ∗ , (·) T, and (·) H, respectively. The trace and determinan… view at source ↗
Figure 2
Figure 2. Illustration of the FG designed for the proposed MMSE algorithm. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the tripartite FG for bilinear inference of [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Root mean square error (RMSE) of γˆ for different antenna configurations. TABLE I SIMULATION PARAMETERS Parameter Symbol Value Forward gain of the repeater |α| 2 [dB] 10 Reverse gain of the repeater |β| 2 [dB] 10 Num. of iterations for A, B estimation NIter 100 Num. of…
Figure 5
Figure 5. Figure 5: Iterative convergence behavior of NLS-based and MMSE-based [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 3 linked inside Pith

  1. [27]

    Reciprocity calibration of dual-antenna repeaters,

    E. G. Larsson, J. Vieira, and P. Frenger, “Reciprocity calibration of dual-antenna repeaters,”IEEE Wireless Commun. Lett., vol. 13, no. 6, pp. 1606–1610, 2024

    2024

  2. [1]

    Noncooperative cellular wireless with unlimited num- bers of base station antennas,

    T. L. Marzetta, “Noncooperative cellular wireless with unlimited num- bers of base station antennas,”IEEE Trans. Wireless Commun., vol. 9, no. 11, pp. 3590–3600, 2010

    2010

  3. [2]

    Massive MIMO for next generation wireless systems,

    E. G. Larsson, O. Edfors, F. Tufvesson, and T. L. Marzetta, “Massive MIMO for next generation wireless systems,”IEEE Commun. Mag., vol. 52, no. 2, pp. 186–195, 2014

    2014

  4. [3]

    Non-reciprocal transceivers in OFDM/SDMA systems: Impact and mitigation,

    A. Bourdoux, B. Come, and N. Khaled, “Non-reciprocal transceivers in OFDM/SDMA systems: Impact and mitigation,” inProc. IEEE Radio Wireless Conf. (RAWCON), Boston, MA, USA, 2003

    2003

  5. [4]

    Massive MIMO systems with non-ideal hardware: Energy efficiency, estimation, and capacity limits,

    E. Bj ¨ornson, J. Hoydis, M. Kountouris, and M. Debbah, “Massive MIMO systems with non-ideal hardware: Energy efficiency, estimation, and capacity limits,”IEEE Trans. Inf. Theory, 2014

    2014

  6. [5]

    Relative channel reciprocity calibration in MIMO/TDD systems,

    F. Kaltenberger, D. Gesbert, and R. Knopp, “Relative channel reciprocity calibration in MIMO/TDD systems,” inProc. Future Netw. Mobile Summit, Florence, Italy, Jun. 2010

    2010

  7. [6]

    Reciprocity calibration for massive MIMO: Proposal, modeling, and validation,

    J. Vieira, F. Rusek, O. Edfors, S. Malkowsky, L. Liu, and F. Tufvesson, “Reciprocity calibration for massive MIMO: Proposal, modeling, and validation,”IEEE Trans. Wireless Commun., 2017

    2017

  8. [7]

    A framework for over-the-air reciprocity calibration for TDD massive MIMO systems,

    X. Jiang, A. Decurninge, K. Gopala, F. Kaltenberger, M. Guillaud, and D. Slock, “A framework for over-the-air reciprocity calibration for TDD massive MIMO systems,”IEEE Trans. Wireless Commun., pp. 5975– 5990, Sept. 2018

    2018

Show all 36 references
  1. [8]

    Cell-free massive MIMO versus small cells,

    H. Q. Ngo, A. Ashikhmin, H. Yang, E. G. Larsson, and T. L. Marzetta, “Cell-free massive MIMO versus small cells,”IEEE Trans. Wireless Commun., vol. 16, no. 3, pp. 1834–1850, 2017

    2017

  2. [9]

    Making cell-free massive MIMO competitive with MMSE processing and centralized implementation,

    E. Bj ¨ornson and L. Sanguinetti, “Making cell-free massive MIMO competitive with MMSE processing and centralized implementation,” IEEE Trans. Wireless Commun., vol. 19, no. 1, pp. 77–90, 2020

    2020

  3. [10]

    Performance analysis of cell-free massive MIMO systems: A stochastic geometry approach,

    A. Papazafeiropoulos, P. Kourtessis, M. D. Renzo, S. Chatzinotas, and J. M. Senior, “Performance analysis of cell-free massive MIMO systems: A stochastic geometry approach,”IEEE Trans. Veh. Technology, vol. 69, no. 4, pp. 3523–3537, 2020

    2020

  4. [11]

    Grant-free access via bilinear inference for cell-free MIMO with low- coherence pilots,

    H. Iimori, T. Takahashi, K. Ishibashi, G. T. F. de Abreu, and W. Yu, “Grant-free access via bilinear inference for cell-free MIMO with low- coherence pilots,”IEEE Trans. Wireless Commun., vol. 20, no. 11, pp. 7694–7710, Nov. 2021

    2021

  5. [12]

    Bayesian receiver design via bilinear inference for cell- free massive MIMO with low-resolution ADCs,

    T. Takahashi, H. Iimori, K. Ando, K. Ishibashi, S. Ibi, and G. T. F. de Abreu, “Bayesian receiver design via bilinear inference for cell- free massive MIMO with low-resolution ADCs,”IEEE Trans. Wireless Commun., vol. 22, no. 7, pp. 4756–4772, 2023

    2023

  6. [13]

    User-centric cell-free massive MIMO networks: A survey of opportunities, challenges and solutions,

    H. A. Ammar, R. Adve, S. Shahbazpanahi, G. Boudreau, and K. V . Srinivas, “User-centric cell-free massive MIMO networks: A survey of opportunities, challenges and solutions,”IEEE Commun. Surv. Tutor., vol. 24, no. 1, pp. 611–652, 2022

    2022

  7. [14]

    Achieving distributed MIMO performance with repeater- assisted cellular massive MIMO,

    S. Willhammar, H. Iimori, J. Vieira, L. Sundstr ¨om, F. Tufvesson, and E. G. Larsson, “Achieving distributed MIMO performance with repeater- assisted cellular massive MIMO,”IEEE Commun. Mag., vol. 63, no. 3, pp. 114–119, Mar. 2025

    2025

  8. [15]

    Study on reconfigurable repeater- based RSMA systems,

    T.-H. Vu, N. H. Tu, and V . N. Q. Bao, “Study on reconfigurable repeater- based RSMA systems,”IEEE Wireless Commun. Lett., vol. 14, no. 5, pp. 1371–1375, May 2025

    2025

  9. [17]

    Amplification strategy in repeater-assisted MIMO systems via mi- norization maximization,

    H. Iimori, E. Kurihara, T. Yoshida, J. Vieira, and S. Malomsoky, “Amplification strategy in repeater-assisted MIMO systems via mi- norization maximization,” inProc. IEEE Global Communications Conf. (GLOBECOM), Kuala Lumpur, Malaysia, Dec. 2023

    2023

  10. [18]

    Is repeater-assisted massive MIMO compatible with dynamic TDD?

    M. Andersson, A. Chowdhury, and E. G. Larsson, “Is repeater-assisted massive MIMO compatible with dynamic TDD?” arXiv preprint arXiv:2510.20998, Jan. 2026. [Online]. Available: https://arxiv.org/abs/2510.20998

    2026 arXiv

  11. [19]

    Fair and energy-efficient activation control mechanisms for repeater-assisted massive MIMO,

    O. A. Topal, O. T. Demir, E. Bj ¨ornson, and C. Cavdar, “Fair and energy-efficient activation control mechanisms for repeater-assisted massive MIMO,”arXiv preprint arXiv:2504.03428, Apr. 2025. [Online]. Available: https://arxiv.org/abs/2504.03428

    2025 arXiv

  12. [20]

    On the performance of dual-antenna repeater assisted bi-static MIMO ISAC,

    A. Chowdhury and E. G. Larsson, “On the performance of dual-antenna repeater assisted bi-static MIMO ISAC,” arXiv preprint arXiv:2511.17980, Jan. 2026. [Online]. Available: https://arxiv.org/abs/2511.17980

    2026

  13. [21]

    Repeater swarm- assisted cellular systems: Interaction stability and performance analysis,

    J. Bai, A. Chowdhury, A. Hansson, and E. G. Larsson, “Repeater swarm- assisted cellular systems: Interaction stability and performance analysis,” IEEE Trans. Wireless Commun., vol. 25, pp. 10 018–10 034, Jan. 2026

    2026

  14. [22]

    Active-reconfigurable-repeater-assisted NOMA networks in internet of things: Reliability, security, and covertness,

    A.-T. Le, T.-H. Vu, N. H. Tu, T. N. Nguyen, L.-T. Tu, and M. V oznak, “Active-reconfigurable-repeater-assisted NOMA networks in internet of things: Reliability, security, and covertness,”IEEE Internet Things J., vol. 12, no. 7, pp. 8759–8772, Apr. 2025

    2025

  15. [23]

    Stability analysis of interacting wireless repeaters,

    E. G. Larsson and J. Bai, “Stability analysis of interacting wireless repeaters,” inProc. IEEE Int. Workshop Signal Process. Adv. Wireless Commun. (SPAWC), Lucca, Italy, Sept. 2024

    2024

  16. [24]

    Network-controlled repeater – An introduction,

    F. I. G. Carvalho, R. V . de O. Paiva, T. F. Maciel, V . F. Monteiro, F. R. M. Lima, D. C. Moreira, D. A. Sousa, B. Makki, M. ˚Astr¨om, and L. Bao, “Network-controlled repeater – An introduction,”IEEE Commun. Stand. Mag., Aug. 2025, early Access

    2025

  17. [25]

    Relaying systems with reciprocity mismatch: Impact analysis and calibration,

    R. Nie, L. Chen, N. Zhao, Y . Chen, F. R. Yu, and G. Wei, “Relaying systems with reciprocity mismatch: Impact analysis and calibration,” IEEE Trans. Commun., vol. 68, no. 7, pp. 4035–4049, 2020

    2020

  18. [26]

    Channel estimation error and beam- forming performance in repeater-enhanced massive MIMO systems,

    Y . Ma, D. Zhu, B. Li, and P. Liang, “Channel estimation error and beam- forming performance in repeater-enhanced massive MIMO systems,” in Proc. IEEE 26th Annu. Int. Symp. Pers., Indoor, Mobile Radio Commun. (PIMRC), Hong Kong, China, Aug. 2015

    2015

  19. [28]

    Bayesian joint channel and data estimation for correlated large MIMO with non-orthogonal pilots,

    K. Ito, T. Takahashi, S. Ibi, and S. Sampei, “Bayesian joint channel and data estimation for correlated large MIMO with non-orthogonal pilots,” inProc. IEEE Int. Conf. Commun. (ICC), Montreal, Canada, June 2021

    2021

  20. [29]

    Bayesian bilinear inference for joint channel tracking and data detection in millimeter-wave MIMO systems,

    T. Takahashi, H. Iimori, K. Ishibashi, S. Ibi, and G. T. F. de Abreu, “Bayesian bilinear inference for joint channel tracking and data detection in millimeter-wave MIMO systems,”IEEE Trans. Wireless Commun., vol. 23, no. 9, pp. 11 136–11 153, 2024

    2024

  21. [30]

    Joint channel, data, and radar parameter estimation for AFDM systems in doubly-dispersive channels,

    K. R. R. Ranasinghe, H. Seok Rou, G. Thadeu Freitas de Abreu, T. Takahashi, and K. Ito, “Joint channel, data, and radar parameter estimation for AFDM systems in doubly-dispersive channels,”IEEE Trans. Wireless Commun., vol. 24, no. 2, pp. 1602–1619, 2025

    2025

  22. [31]

    On the impact of hardware impairments on massive MIMO,

    U. Gustavsson, C. Sanch ´ez-Perez, T. Eriksson, F. Athley, G. Durisi, P. Landin, K. Hausmair, C. Fager, and L. Svensson, “On the impact of hardware impairments on massive MIMO,” inProc. IEEE Glob. Commun. Conf. Workshops (GLOBECOM WS), 2014

    2014

  23. [32]

    Analysis of massive MIMO with hardware impairments and different channel models,

    F. Athley, G. Durisi, and U. Gustavsson, “Analysis of massive MIMO with hardware impairments and different channel models,” inProc. Eur. Conf. Antennas Propag. (EuCAP), Lisbon, Portugal, 2015

    2015

  24. [33]

    Bilinear gaussian belief propagation for massive MIMO detection with non-orthogonal pilots,

    K. Ito, T. Takahashi, S. Ibi, and S. Sampei, “Bilinear gaussian belief propagation for massive MIMO detection with non-orthogonal pilots,” IEEE Trans. Commun., vol. 72, no. 2, pp. 1045–1061, 2024

    2024

  25. [34]

    Joint activity and channel estimation for extra-large MIMO systems,

    H. Iimori, T. Takahashi, K. Ishibashi, G. T. F. de Abreu, D. Gonz ´alez G., and O. Gonsa, “Joint activity and channel estimation for extra-large MIMO systems,”IEEE Trans. Wireless Commun., vol. 21, no. 9, pp. 7253–7270, 2022

    2022

  26. [35]

    Chockalingam and B

    A. Chockalingam and B. S. Rajan,Large MIMO Systems. Cambridge University Press, 2014

    2014

  27. [36]

    Reciprocity calibration methods for massive MIMO based on antenna coupling,

    J. Vieira, F. Rusek, and F. Tufvesson, “Reciprocity calibration methods for massive MIMO based on antenna coupling,” inProc. IEEE Glob. Commun. Conf. (GLOBECOM), Austin, USA, Dec. 2014

    2014

  28. [2025]

    Available: https://arxiv.org/abs/2509.19119

    [Online]. Available: https://arxiv.org/abs/2509.19119

    arXiv

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.