Pith. sign in

REVIEW 4 major objections 5 minor 46 references

Bayesian Sensing for Time-Varying Channels in ISAC Systems

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read The paper claims a two-layer variational Bayesian estimator can recover delay, integer Doppler, and fractional Doppler simultaneously from the re-aligned frequency-domain channel matrix, avoiding the integer–fractional Doppler mismatch of…

desk verdict Real algorithmic idea in a crowded subfield, but the central claim is validated only under a perfect-channel-estimate assumption the paper promises to relax and never does. read the letter →

arxiv 2504.15042 v1 pith:5UXMV2TZ submitted 2025-04-21 eess.SP

Xueyang Wang , Kai Wu , J. Andrew Zhang , Shiqi Gong , Chengwen Xing This is my paper · ORCID
classification eess.SP
keywords integratedsensingandcommunicationtime-varyingchannelssparseBayesianlearningvariationalinferenceDopplerestimationdelaymultiplemeasurementsignalrecoveryCramér-Raobound
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 sets out to show that in a fast-fading OFDM-based ISAC system, the delay, integer Doppler, and fractional Doppler of multiple targets can be recovered at once from the re-aligned frequency-domain channel matrix by solving a three-dimensional multiple-measurement sparse-signal-recovery problem. The proposed two-layer variational Bayesian inference alternates between a Doppler layer and a delay layer, so the three parameters are read from the same nonzero entries of a sparse tensor and are matched by construction. A simplified two-stage variant first estimates delays with MUSIC and then estimates Doppler with VBI at lower complexity. The authors derive the Cramér-Rao bound for these parameters and their simulations show the two-layer estimator approaching that bound and outperforming two-stage and OTFS-based benchmarks. If this is right, ISAC receivers in high-mobility links could sense targets without a separate Doppler-association step.

What carries the argument

The load-bearing object is the re-aligned frequency-domain channel matrix $\tilde{\mathbf H}_{fd}$, formed by circularly shifting the $m$-th column of the estimated channel matrix $\hat{\mathbf H}_{fd}$ upward by $(m-1)$ positions so that each row corresponds to one quantized Doppler shift. Its factored form $\tilde{\mathbf H}_{fd} = \mathbf A_\tau \mathcal{D}(\mathbf I_N \otimes \mathbf A_\nu^H) + \tilde{\mathbf W}_{fd}$ recasts sensing as a 3D multiple-measurement sparse-signal-recovery problem, where $\mathcal{D} \in \mathbb{C}^{P\times Q\times N}$ has nonzero entries only at delay-fractional-Doppler-integer-Doppler grid points. The two-layer variational Bayesian inference is the mechanism that solves this problem without Kronecker-style vectorization: a first layer estimates the intermediate row-sparse matrix $\mathbf C(n)=\mathbf X(n)\mathbf A_\tau^H$, and a second layer estimates $\mathbf D(n)$ and its precision matrix $\Gamma_d(n)$, whose column statistics feed back into the precision $\Gamma_c(n)$ of the first layer. This feedback loop couples the integer Doppler, fractional Doppler, and delay estimates in one solution.

What would settle it

Run the same Doppler MSE experiment but feed $\hat{\mathbf H}_{fd}$ from an actual least-squares channel estimator (for example, with known pilot symbols) instead of the perfect matrix assumed in Section II-B; if the two-layer VBI's MSE no longer approaches the Cramér-Rao bound at high SNR, the paper's central performance claim depends on an unmodeled error source.

Watch

Extended reading notes

Core claim

At the center of the paper is the claim that the frequency-domain channel matrix of a time-varying OFDM link, after a column-wise circular realignment, has the structure $\tilde{\mathbf H}_{fd}(n) = \mathbf A_\tau \mathbf D(n) \mathbf A_\nu^H + \tilde{\mathbf W}_{fd}(n)$, in which the sparse tensor $\mathcal{D}$ carries all sensing information. Estimating this tensor solves the sensing problem: the row, column, and slice indices of its nonzero entries are the delay, fractional Doppler, and integer Doppler of the targets, so the three parameters come out matched by construction. The proposed two-layer variational Bayesian inference estimates the sparse tensor without vectorizing the full 3D dictionary; a Doppler layer estimates an intermediate row-sparse matrix and a delay layer refines the precision matrix that the Doppler layer needs. A simplified two-stage variant estimates delays by MUSIC first and then uses VBI on a smaller dictionary for the Doppler, trading a little accuracy for lower complexity. The paper also derives the Cramér-Rao bound for $\tau_l$, $\nu_l$, and the complex path coefficients, and reports Doppler mean-square error close to that bound in simulation.

Load-bearing premise

The load-bearing premise is that the re-aligned frequency-domain channel matrix $\hat{\mathbf H}_{fd}$ is already known accurately: the paper explicitly assumes this in Section II-B and promises but never delivers an analysis of the impact of channel estimation error, so an inaccurate channel estimate would flow unmodeled into the sparse recovery, the Doppler and delay estimates, and the CRB comparison.

Editorial extensions

If this is right

  • Multi-target sensing in fast-fading OFDM is reduced to a single sparse recovery problem, so integer Doppler, fractional Doppler, and delay are associated by construction rather than by a post-hoc matching step.
  • Because the two-layer scheme avoids Kronecker-style vectorization of the full 3D dictionary, its complexity is substantially lower than a direct large sparse-recovery solution, at the price of a nested iteration.
  • The MUSIC-based variant cuts the iteration count by estimating delays non-iteratively and reduces the Doppler dictionary dimension from $N$ to $L$, which is suited to low-complexity receivers.
  • The derived Cramér-Rao bound gives a target for any frequency-domain time-varying channel estimator; the reported MSE close to the CRB suggests the realignment preserves the information needed for sensing.
  • The method remains effective over a range of velocities up to normalized Doppler $2.8f_0$ and across different target counts, supporting the paper's claim of wide applicability.

Reading between the lines

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

  • Editorial inference: if the alternating precision-feedback idea is waveform-agnostic, it may transfer to joint angle-delay-Doppler estimation by replacing the delay dictionary with a joint angle-delay dictionary.
  • Editorial inference: the practical accuracy ceiling is likely set by the quality of $\hat{\mathbf H}_{fd}$; a natural extension is to feed the sensing stage with an explicit channel estimator and quantify how the gap to the CRB grows.
  • Editorial inference: the stacking construction behind the MUSIC stage may also resolve ambiguities in other multiple-measurement MUSIC settings where summed correlation matrices smear closely spaced targets; the paper does not claim this generalization.
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 / 5 minor

Summary. The manuscript proposes a frequency-domain sensing method for OFDM-based ISAC under fast time-varying channels. It formulates joint delay, integer Doppler, and fractional Doppler estimation as a 3D multiple-measurement sparse signal recovery problem using the re-aligned channel matrix in Eq. (7). The main contribution is a two-layer variational Bayesian inference (VBI) algorithm that alternates between estimating the row-sparse matrix C(n) and the sparse matrix D(n), thereby avoiding the parameter mismatch of separately estimated integer and fractional Dopplers. A lower-complexity two-stage MUSIC-based VBI variant is also proposed, and a Cramér-Rao bound is derived for the sensing parameters. Simulation results compare the MSE of the proposed methods against FFT, OTFS-based, two-stage VBI, and EM-VB baselines, and show the two-layer VBI approaching the perfect-VBI benchmark and the CRB.

Significance. The contribution is potentially useful: estimating integer and fractional Doppler jointly in the same sparse Bayesian framework is a genuine step beyond the two-step estimation in [31], the two-layer decomposition is an interesting way to avoid vectorizing a 3D dictionary, and the factor-graph presentation with appendices makes the algorithm structure transparent. The paper also provides concrete MSE comparisons over SNR, target number, and velocity, and the simplified MUSIC-VBI variant offers a clear complexity-performance trade-off. The CRB derivation from the frequency-domain input-output model is a useful reference point, although it is not directly the bound for the estimator's actual input. Overall, if the technical errors below are fixed and the missing error analysis is supplied, the work would be a solid contribution to ISAC sensing in time-varying channels.

major comments (4)
  1. [II-B and Section V] Section II-B explicitly states that the estimated channel matrix H_hat_fd is assumed known and that the impact of its estimation error 'will be discussed', but no such discussion appears in Sections III-VII. The entire sensing pipeline (Algorithms 1 and 2) consumes the re-aligned channel matrix tilde_H_fd of Eq. (7) as a direct noisy observation, and the CRB in Section V is derived from r(k)=H_fd(k)s(k)+w_1(k) in Eq. (55), not from an estimated channel matrix. Consequently, the claimed 'approach the perfect-VBI and CRB' behavior reported in Section VI-A is demonstrated only for an idealized input. A realistic channel-estimation error, especially correlated or non-Gaussian error, is neither characterized nor propagated. I request either an explicit error analysis for the sensing estimators under channel-estimation error, an end-to-end simulation that estimates H_hat_fd from pilots under the same ICI, or a revised claim that restricts the results to the case where the channel matrix is available.
  2. [III-D, Lemma 1 and Appendix B] The derivation of q(alpha) in Eq. (75) contains '+ N ln p(alpha)', which multiplies the Gamma prior by N and is inconsistent with the joint distribution in Eq. (28), where p(alpha) appears once. This leads to the posterior shape parameter a+MNQ in Lemma 1; the shape parameter should reflect the number of scalar observations in the likelihood, and the displayed formula introduces a Q-dependent term that is not present in the likelihood. The same error pattern appears in Lemma 3, where the shape is written as a+N P Q. Since these shape parameters enter every VBI iteration, the update equations as printed are not correct as written. Please correct the derivations or justify the additional factors.
  3. [IV, Eq. (54)] The pseudo-inverse in Eq. (54) is written as hat_A_tau^dagger = hat_A_tau (hat_A_tau^H hat_A_tau), which is dimensionally N by L and misses the inverse. Right-multiplying Eq. (53) to remove hat_A_tau^H requires the right inverse (hat_A_tau^H)^dagger = hat_A_tau (hat_A_tau^H hat_A_tau)^{-1}. As written, Eq. (54) does not follow from Eq. (53), and the derivation of the simplified two-stage MUSIC-VBI method is therefore incomplete.
  4. [VI and Algorithm 1] The delay grid size P, the Doppler grid size Q, the grid spacings, and the Gamma hyperparameters a, b, c, d, e, f of Section III-B are never specified in Section VI. Without these values the MM-SSR formulation, the complexity expressions, and the simulation results cannot be reproduced. In addition, Algorithm 1, Step 12, computes the outer-loop convergence test by dividing a change in Gamma_c by ||Gamma_x2(n)||^2, mixing two different precision matrices; this appears to be a typo and should be corrected to a consistent norm.
minor comments (5)
  1. [VI-A, Fig. 6 discussion] The sentence 'It is evident that the MSE degrades as the SNR increases' states the opposite of the expected and plotted trend; it should say that the MSE improves (decreases) as the SNR increases.
  2. [Algorithm 1, Step 12] The convergence criterion divides the Gamma_c residual by ||Gamma_x2(n)||^2, which is inconsistent; the denominator should involve the same matrix as the numerator.
  3. [IV, Eq. (50)] The noise term in Eq. (50) is written as sigma^2 I_M, but M is not defined; from context it should be sigma^2 I_N.
  4. [III-B and III-D] The symbol M is used for the number of columns of C(n) and Y(n), which equals N in this model; this conflicts with the use of N as the number of subcarriers and should be clarified or renamed.
  5. [References] Reference [45] is incomplete: it lacks the article title, volume, and page numbers, which prevents the reader from locating it.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the two-layer VBI and MUSIC-based estimators are derived from the signal model and checked against an independently computed CRB and external benchmarks; the unfulfilled channel-estimation-error analysis is a completeness gap, not a circular step.

full rationale

Walked the derivation chain. The sensing problem is set up from the re-aligned frequency-domain channel matrix ~H_fd in Eq. (7), which is modeled as a grid dictionary A_tau, A_nu acting on a 3D sparse matrix D (Eqs. (8)-(10)). The two-layer VBI (Algorithm 1) estimates D by alternating updates (32)-(36) and the supporting lemmas; delays and Dopplers are read off the support of D. Nothing in this chain defines the target parameters in terms of the estimator output, and no fitted parameter is later renamed as a prediction. The CRB in Section V is computed by standard Fisher information from the received-signal model r(k) = H_fd(k)s(k) + w1(k) (Eq. (55)), not from the VBI updates, so matching it in Figs. 3-7 is an external check, not a tautology. Simulations compare against FFT [31], OTFS sensing [44], two-stage VBI [38], EM-VB [45], and perfect-VBI [38], which do not depend on the proposed method. Self-citations: Eq. (1) adopts the channel model from [31] and the rectangular-window assumption is attributed to [31]; since [31] shares authors with this paper, this is a self-citation, but it is a published, parameter-free model used as a premise, not a uniqueness theorem or fitted input, and the paper's contribution is the new VBI/MUSIC estimation machinery. Flagged completeness gap: Section II-B states 'we assume that the estimated channel matrix H_fd is known for developing sensing algorithms, and will discuss the impact of H_fd estimation error on proposed designs,' but no such analysis appears; this weakens the end-to-end validation but is a missing-analysis or scope issue, not circularity, because H_fd is an intermediate input rather than the predicted sensing parameter, and the CRB is not derived from H_fd. Therefore no circular step can be exhibited.

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

The proposed method relies on a standard SBL hierarchical prior, a grid-based dictionary, and a known channel matrix. The main free parameters are algorithmic (grid sizes, hyperparameters, initial values), and several are left unspecified, which weakens reproducibility.

free parameters (6)
  • Delay grid size P
    The dictionary Aτ is C^{N×P} but P is never specified in the paper; the estimator's accuracy and complexity depend on it.
  • Doppler grid size Q
    The dictionary Aν is C^{K×Q} but Q is never specified; Q is required to run the sparse recovery.
  • Grid spacing for delay and Doppler
    The resolution of the delay and Doppler grids is not given; it determines the minimum parameter separation the algorithm can resolve.
  • Gamma hyperparameters a, b, c, d, e, f = not specified
    The shape and inverse-scale parameters of the Gamma priors in (15), (17), (24), (26) are not assigned numerical values.
  • Initial precisions and means = 1
    Initial values in Algorithm 1 are set to 1 for α, β, γ_d, γ_c, and u, as stated in Section VI.
  • Number of targets L
    L is used to construct the MUSIC null-space and to select peaks; it is assumed known rather than estimated.
assumptions (6)
  • domain assumption The time-varying channel is composed of L discrete specular paths with constant delay and Doppler over the observation window, as modeled in (1).
    Used throughout Section II; standard for ISAC but an idealization of real channels.
  • domain assumption An accurate estimate of the channel matrix Ĥ_fd is available.
    Section II-B states this assumption and promises a discussion of its impact, which is absent.
  • domain assumption The true delay and Doppler values lie on the predefined grids.
    The SSR formulation in (8) uses grid dictionaries Aτ and Aν; off-grid parameters cause basis mismatch, which is not analyzed.
  • domain assumption The number of targets L is known.
    Used in (51) to build UN and to extract peaks; no estimation of L is provided.
  • domain assumption No clock asynchrony between transmitter and sensing receiver.
    Stated in Section II; excludes a practical synchronization impairment.
  • standard math Mean-field factorization of the variational posterior.
    Standard VBI assumption, used in Appendix A to derive the update equations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bayesian Sensing for Time-Varying Channels in ISAC Systems." pith.science (2026). https://pith.science/paper/5UXMV2TZ

@misc{pith2026250415042,
  author       = {Pith},
  title        = {Pith review of: Bayesian Sensing for Time-Varying Channels in ISAC Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5UXMV2TZ}},
  note         = {Machine review of arXiv:2504.15042}
}
read the original abstract

Future mobile networks are projected to support integrated sensing and communications in high-speed communication scenarios. Nevertheless, large Doppler shifts induced by time-varying channels may cause severe inter-carrier interference (ICI). Frequency domain shows the potential of reducing ISAC complexity as compared with other domains. However, parameter mismatching issue still exists for such sensing. In this paper, we develop a novel sensing scheme based on sparse Bayesian framework, where the delay and Doppler estimation problem in time-varying channels is formulated as a 3D multiple measurement-sparse signal recovery (MM-SSR) problem. We then propose a novel two-layer variational Bayesian inference (VBI) method to decompose the 3D MM-SSR problem into two layers and estimate the Doppler in the first layer and the delay in the second layer alternatively. Subsequently, as is benefited from newly unveiled signal construction, a simplified two-stage multiple signal classification (MUSIC)-based VBI method is proposed, where the delay and the Doppler are estimated by MUSIC and VBI, respectively. Additionally, the Cram\'er-Rao bound (CRB) of the considered sensing parameters is derived to characterize the lower bound for the proposed estimators. Corroborated by extensive simulation results, our proposed method can achieve improved mean square error (MSE) than its conventional counterparts and is robust against the target number and target speed, thereby validating its wide applicability and advantages over prior arts.

Figures

Figures reproduced from arXiv: 2504.15042 by the authors.

Figure 1
Figure 1. The considered ISAC scenario with high-speed target [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The factor graph of the proposed two-layer VBI method [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 5
Figure 5. The MSE performance for Doppler estimation of our pro [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The MSE performance for Doppler estimation with diffe [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 7
Figure 7. Figure 7: The MSE performance for delay estimation with differ [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: The MSE performance for delay estimation of our propo [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 40 canonical work pages

  1. [31]

    Frequency-doma in sensing in time-varying channels,

    Y . Sun, J. A. Zhang, K. Wu, and R. P . Liu, “Frequency-doma in sensing in time-varying channels,” IEEE Wireless Commun. Lett. , vol. 12, no. 1, pp. 16–20, Jan. 2023

    work page 2023

  2. [1]

    Joint communications and sensing employing optimized MIM O-OFDM signals,

    K. Wu, J. A. Zhang, Z. Ni, X. Huang, Y . J. Guo, and S. Chen, “Joint communications and sensing employing optimized MIM O-OFDM signals,” IEEE Internet Things J. , vol. 11, no. 6, pp. 10 368–10 383, March 2024

    work page 2024

  3. [2]

    Joint MIMO communications and sensing with hybrid beamfor ming architecture and OFDM waveform optimization,

    S. D. Liyanaarachchi, T. Riihonen, C. B. Barneto, and M. V alkama, “Joint MIMO communications and sensing with hybrid beamfor ming architecture and OFDM waveform optimization,” IEEE Trans. Wireless Commun., vol. 23, no. 2, pp. 1565–1580, Feb. 2024

    work page 2024

  4. [3]

    Code-divi sion OFDM joint communication and sensing system for 6G machine- type communication,

    X. Chen, Z. Feng, Z. Wei, P . Zhang, and X. Y uan, “Code-divi sion OFDM joint communication and sensing system for 6G machine- type communication,” IEEE Internet Things J. , vol. 8, no. 15, pp. 12 093– 12 105, Aug. 2021

    work page 2021

  5. [4]

    Joint radar and communication design: Applications, state-of-t he-art, and the road ahead,

    F. Liu, C. Masouros, A. P . Petropulu, H. Griffiths, and L. H anzo, “Joint radar and communication design: Applications, state-of-t he-art, and the road ahead,” IEEE Trans. Commun. , vol. 68, no. 6, pp. 3834–3862, Jun. 2020

    2020

  6. [5]

    An overview of signal processing techniques for joint communication and radar sensing,

    J. A. Zhang, F. Liu, C. Masouros, R. W. Heath, Z. Feng, L. Zh eng, and A. Petropulu, “An overview of signal processing techniques for joint communication and radar sensing,” IEEE J. Sel. Topics Signal Process. , vol. 15, no. 6, pp. 1295–1315, Nov. 2021

    2021

  7. [6]

    To ward multi- functional 6G wireless networks: Integrating sensing, com munication, and security,

    Z. Wei, F. Liu, C. Masouros, N. Su, and A. P . Petropulu, “To ward multi- functional 6G wireless networks: Integrating sensing, com munication, and security,” IEEE Commun. Mag. , vol. 60, no. 4, pp. 65–71, April 2022

    work page 2022

  8. [7]

    Integrated sensing and com munication for symbiotic radio with multiple IoT devices,

    Q. Tao, C. Huang, and X. Chen, “Integrated sensing and com munication for symbiotic radio with multiple IoT devices,” IEEE Commun. Lett. , vol. 28, no. 8, pp. 1820–1824, Aug. 2024

    work page 2024

Show all 46 references
  1. [8]

    Two- stage multiband Wi-Fi sensing for ISAC via stochastic parti cle-based variational bayesian inference,

    Z. Hu, A. Liu, Y . Wan, T. Q. S. Quek, and M.-J. Zhao, “Two- stage multiband Wi-Fi sensing for ISAC via stochastic parti cle-based variational bayesian inference,” in GLOBECOM 2023 - 2023 IEEE Global Communications Conference , 2023, pp. 5617–5622

    2023

  2. [9]

    Bayesian learning for double-RIS ai ded ISAC systems with superimposed pilots and data,

    X. Gan, C. Huang, Z. Y ang, C. Zhong, X. Chen, Z. Zhang, Q. Gu o, C. Y uen, and M. Debbah, “Bayesian learning for double-RIS ai ded ISAC systems with superimposed pilots and data,” IEEE J. Sel. Topics Signal Process., vol. 18, no. 5, pp. 766–781, July 2024

    2024

  3. [10]

    Joint sparse bayesian learning for ch annel esti- mation in ISAC,

    K. Chen and C. Qi, “Joint sparse bayesian learning for ch annel esti- mation in ISAC,” IEEE Commun. Lett. , vol. 28, no. 8, pp. 1825–1829, Aug. 2024

    2024

  4. [11]

    Enabling joint communication and radar sensing in mobile networks—a survey,

    J. A. Zhang, M. L. Rahman, K. Wu, X. Huang, Y . J. Guo, S. Che n, and J. Y uan, “Enabling joint communication and radar sensing in mobile networks—a survey,” IEEE Commun. Surveys and Tutorials , vol. 24, no. 1, pp. 306–345, Oct. 2022

    2022

  5. [12]

    Bounding performance and su ppressing intercarrier interference in wireless mobile OFDM,

    X. Cai and G. B. Giannakis, “Bounding performance and su ppressing intercarrier interference in wireless mobile OFDM,” IEEE Trans. Com- mun., vol. 51, no. 12, pp. 2047–2056, 2003

    2003

  6. [13]

    Y . S. Cho, J. Kim, W. Y . Y ang, and C. G. Kang, MIMO-OFDM Wireless Communications With MATLAB . Singapore: Wiley (Asia) Pte. Ltd., 2010

    2010

  7. [14]

    Transmission tec hniques for digital terrestrial TV broadcasting,

    H. Sari, G. Karam, and I. Jeanclaude, “Transmission tec hniques for digital terrestrial TV broadcasting,” IEEE Commun. Mag. , vol. 33, no. 2, pp. 100–109, 1995

    1995

  8. [15]

    Spa rse prior-guided deep learning for OTFS channel estimation,

    X. Zhang, C. Liu, W. Y uan, J. A. Zhang, and D. W. K. Ng, “Spa rse prior-guided deep learning for OTFS channel estimation,” IEEE Trans. V eh. Technol., vol. 73, no. 12, pp. 19 913–19 918, Dec. 2024

    2024

  9. [16]

    OTFS: a new generation of modulat ion addressing the challenges of 5G,

    R. Hadani and A. Monk, “OTFS: a new generation of modulat ion addressing the challenges of 5G,” arXiv preprint arXiv:1802.02623 , 2018

    2018 arXiv

  10. [17]

    Orthogonal time frequency spa ce modu- lation,

    R. Hadani, S. Rakib, M. Tsatsanis, A. Monk, A. J. Goldsmi th, A. F. Molisch, and R. Calderbank, “Orthogonal time frequency spa ce modu- lation,” in Proc. IEEE WCNC , 2017, pp. 1–6

    2017

  11. [18]

    Orthogonal time-frequency space modulation: A p romising next-generation waveform,

    Z. Wei, W. Y uan, S. Li, J. Y uan, G. Bharatula, R. Hadani, a nd L. Hanzo, “Orthogonal time-frequency space modulation: A p romising next-generation waveform,” IEEE Wireless Commun., vol. 28, no. 4, pp. 136–144, Aug. 2021

    2021

  12. [19]

    Orthogo nal time frequency space modulation—part III: ISAC and potential ap plications,

    W. Y uan, Z. Wei, S. Li, R. Schober, and G. Caire, “Orthogo nal time frequency space modulation—part III: ISAC and potential ap plications,” IEEE Commun. Lett. , vol. 27, no. 1, pp. 14–18, Jan. 2023

    2023

  13. [20]

    Integrate d sensing and communication-assisted orthogonal time frequency spa ce transmis- sion for vehicular networks,

    W. Y uan, Z. Wei, S. Li, J. Y uan, and D. W. K. Ng, “Integrate d sensing and communication-assisted orthogonal time frequency spa ce transmis- sion for vehicular networks,” IEEE J. Sel. Topics Signal Process., vol. 15, no. 6, pp. 1515–1528, Nov. 2021

    2021

  14. [21]

    Robust N OMA- assisted OTFS-ISAC network design with 3-D motion predicti on topol- ogy,

    L. Xiang, K. Xu, J. Hu, C. Masouros, and K. Y ang, “Robust N OMA- assisted OTFS-ISAC network design with 3-D motion predicti on topol- ogy,” IEEE Internet Things J. , vol. 11, no. 9, pp. 15 909–15 918, May 2024

    2024

  15. [22]

    Sensing aided uplink transmission in OTFS ISAC with joint p arameter association, channel estimation and signal detection,

    X. Y ang, H. Li, Q. Guo, J. A. Zhang, X. Huang, and Z. Cheng, “Sensing aided uplink transmission in OTFS ISAC with joint p arameter association, channel estimation and signal detection,” IEEE Trans. V eh. Technol., vol. 73, no. 6, pp. 9109–9114, Jun. 2024

    2024

  16. [23]

    Off-grid c hannel estimation with sparse Bayesian learning for OTFS systems,

    Z. Wei, W. Y uan, S. Li, J. Y uan, and D. W. K. Ng, “Off-grid c hannel estimation with sparse Bayesian learning for OTFS systems, ” IEEE Trans. Wireless Commun. , vol. 21, no. 9, pp. 7407–7426, Sep. 2022

    2022

  17. [24]

    Off-grid channel estimation using grid evolution for OTFS systems,

    Y . Shan, F. Wang, Y . Hao, J. Y uan, J. Hua, and Y . Xin, “Off-grid channel estimation using grid evolution for OTFS systems,” IEEE Trans. Wireless Commun., pp. 1–1, 2024

    2024

  18. [25]

    A novel ISAC transmission framework based on spatially-spre ad orthog- onal time frequency space modulation,

    S. Li, W. Y uan, C. Liu, Z. Wei, J. Y uan, B. Bai, and D. W. K. N g, “A novel ISAC transmission framework based on spatially-spre ad orthog- onal time frequency space modulation,” IEEE J. Sel. Areas Commun. , vol. 40, no. 6, pp. 1854–1872, Jun. 2022

    2022

  19. [26]

    DFT-spread orthogonal time f requency space system with superimposed pilots for terahertz integr ated sensing and communication,

    Y . Wu, C. Han, and Z. Chen, “DFT-spread orthogonal time f requency space system with superimposed pilots for terahertz integr ated sensing and communication,” IEEE Trans. Wireless Commun. , vol. 22, no. 11, pp. 7361–7376, Nov. 2023

    2023

  20. [27]

    OTFS-based jo int communication and sensing for future industrial IoT,

    K. Wu, J. A. Zhang, X. Huang, and Y . J. Guo, “OTFS-based jo int communication and sensing for future industrial IoT,” IEEE Internet Things J. , vol. 10, no. 3, pp. 1973–1989, Feb. 2023

    1973

  21. [28]

    Integrated sensing and communicat ion-assisted user state refinement for OTFS systems,

    Y . Shi and Y . Huang, “Integrated sensing and communicat ion-assisted user state refinement for OTFS systems,” IEEE Trans. Wireless Com- mun., vol. 23, no. 2, pp. 922–936, Feb. 2024

    2024

  22. [29]

    Adaptive transmiss ion with frequency-domain precoding and linear equalization over f ast fading channels,

    H. Zhang, X. Huang, and J. A. Zhang, “Adaptive transmiss ion with frequency-domain precoding and linear equalization over f ast fading channels,” IEEE Trans. Wireless Commun. , vol. 20, no. 11, pp. 7420– 7430, Nov. 2021

    2021

  23. [30]

    Low-overhead OTFS transmission with frequency or time domain channel estimation,

    ——, “Low-overhead OTFS transmission with frequency or time domain channel estimation,” IEEE Trans. V eh. Technol., vol. 73, no. 1, pp. 799– 811, Jan. 2024

    2024

  24. [32]

    Sparse Bayesian learning for ba sis selection,

    D. P . Wipf and B. D. Rao, “Sparse Bayesian learning for ba sis selection,” IEEE Trans. Signal Process. , vol. 52, no. 8, pp. 2153–2164, 2004

    2004

  25. [33]

    Sparse Bayesian learning and the releva nce vector machine,

    M. E. Tipping, “Sparse Bayesian learning and the releva nce vector machine,” J. Mach. Learn. Res. , vol. 1, no. Jun., pp. 211–244, 2001

    2001

  26. [34]

    A tutorial on variational Ba yesian inference,

    C. W. Fox and S. J. Roberts, “A tutorial on variational Ba yesian inference,” Artificial Intelligence Rev. , vol. 38, no. 2, pp. 85–95, 2012

    2012

  27. [35]

    Error rate analys is of cogni- tive radio transmissions with imperfect channel sensing,

    G. Ozcan, M. C. Gursoy, and S. Gezici, “Error rate analys is of cogni- tive radio transmissions with imperfect channel sensing,” IEEE Trans. Wireless Commun., vol. 13, no. 3, pp. 1642–1655, March 2014

    2014

  28. [36]

    Sbl-based glrt for spe ctrum sensing in ofdma-based cognitive radio networks,

    A. Mishra and A. K. Jagannatham, “Sbl-based glrt for spe ctrum sensing in ofdma-based cognitive radio networks,” IEEE Commun. Lett. , vol. 20, no. 7, pp. 1433–1436, July 2016

    2016

  29. [37]

    Joint Baye sian channel estimation and data detection for OTFS systems in LE O satellite communications,

    X. Wang, W. Shen, C. Xing, J. An, and L. Hanzo, “Joint Baye sian channel estimation and data detection for OTFS systems in LE O satellite communications,” IEEE Trans. Commun., vol. 70, no. 7, pp. 4386–4399, Jul. 2022

    2022

  30. [38]

    The vari ational approximation for Bayesian inference,

    D. G. Tzikas, A. C. Likas, and N. P . Galatsanos, “The vari ational approximation for Bayesian inference,” IEEE Signal Process. Mag. , vol. 25, no. 6, pp. 131–146, 2008

    2008

  31. [39]

    Direction of departure (DOD) and direction of arrival (DOA) estimation in MIMO radar with reduced- dimension MUSIC,

    X. Zhang, L. Xu, L. Xu, and D. Xu, “Direction of departure (DOD) and direction of arrival (DOA) estimation in MIMO radar with reduced- dimension MUSIC,” IEEE Commun. Lett. , vol. 14, no. 12, pp. 1161– 1163, Dec. 2010

    2010

  32. [40]

    Low-complexity DOA estimat ion based on compressed MUSIC and its performance analysis,

    F. Y an, M. Jin, and X. Qiao, “Low-complexity DOA estimat ion based on compressed MUSIC and its performance analysis,” IEEE Trans. Signal Process., vol. 61, no. 8, pp. 1915–1930, April 2013

    1915

  33. [41]

    Performance ana lysis of an improved MUSIC DoA estimator,

    P . V allet, X. Mestre, and P . Loubaton, “Performance ana lysis of an improved MUSIC DoA estimator,” IEEE Trans. Signal Process., vol. 63, no. 23, pp. 6407–6422, Dec. 2015

    2015

  34. [42]

    Multiple signal classification based joint communication and sensing system,

    X. Chen, Z. Feng, Z. Wei, X. Y uan, P . Zhang, J. A. Zhang, an d H. Y ang, “Multiple signal classification based joint communication and sensing system,” IEEE Trans. Wireless Commun., vol. 22, no. 10, pp. 6504–6517, Oct. 2023

    2023

  35. [43]

    S. M. Kay, Fundamentals of Statistical Signal Processing: Estimatio n Theory. Upper Saddle River, NJ, USA: Prentice Hall, 1993

    1993

  36. [44]

    Radar sensing via OTFS signaling: A delay Doppler signal processi ng perspec- tive,

    K. Zhang, W. Y uan, S. Li, F. Liu, F. Gao, P . Fan, and Y . Cai, “Radar sensing via OTFS signaling: A delay Doppler signal processi ng perspec- tive,” in ICC 2023 - IEEE International Conference on Communications , 2023, pp. 6429–6434

    2023

  37. [45]

    J. Dai, X. Bao, W. Xu, and C. Chang, IEEE Signal Process. Lett

  38. [46]

    Tang, DOA Estimation Based on MUSIC Algorithm , Accessed: Nov

    H. Tang, DOA Estimation Based on MUSIC Algorithm , Accessed: Nov. 28, 2017. [Online]. Available: https://www.divaport al.org/smash/ get/diva2:724272/FULLTEXT01.pdf

    2017

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.