Pith. sign in

REVIEW 3 major objections 5 minor 28 references

One Target, Many Views: Multi-User Fusion for Collaborative Uplink ISAC

T0 review · 3 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 This paper claims that pilot-free multi-user uplink signals can be jointly decoded and used for 3D target sensing, and that fusing many users' views improves both.

desk verdict Worth engaging: the pilot-free multi-user uplink ISAC setup and AoA fusion are useful, but the headline 3D collaborative fusion claim is not supported by the simulations and cuts against the paper's own bistatic model. read the letter →

arxiv 2505.01223 v1 pith:6NAHONRQ submitted 2025-05-02 eess.SP

Sajad Daei , Gabor Fodor , Mikael Skoglund This is my paper · ORCID
classification eess.SP
keywords integratedsensingandcommunicationmulti-useruplinkpilot-freeOFDMatomicnormminimization3Dsuper-resolutiondualpolynomialdelay-Doppler-AoAestimationcollaborativefusion
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 in a millimeter-wave uplink, the data signals users already send can double as radar-like probes, removing the need for dedicated pilot transmissions. The base station treats the channels as sparse sums of delay, Doppler, and angle-of-arrival paths, and solves a 3D super-resolution problem to recover those paths while decoding messages. The central wager is that extra users are not interference but an asset: because each user sees the same target from a different position and with a different velocity, fusing their estimates sharpens target localization and, in turn, message recovery. Numerical experiments with up to five users support this, reporting lower angle-of-arrival error and lower symbol error rate for collaborative fusion at 0 dB signal-to-noise ratio.

What carries the argument

The load-bearing object is the 3D atomic norm and its SDP relaxation: the atomic norm $\|X_i\|_{A_i}$ is the convex proxy for the sparsest sum of atoms $f_i a_{3D}(\zeta)^T$, where $a_{3D}(\zeta)$ is the steering vector spanning subcarrier, OFDM-block, and antenna dimensions. The semidefinite program enforces a three-level Toeplitz structure on the dual variable, which is what allows continuous-valued delays, Dopplers, and AoAs to be recovered without a grid. The companion machinery is the dual polynomial $f_i(\zeta)$, whose peaks mark the estimated paths; the collaborative fusion step aggregates these polynomials across users, and the aligned variant removes user-specific codebook phase offsets before averaging.

What would settle it

Place two users at well-separated distances from the same target so their bistatic delays differ by more than a super-resolution grid cell, run the aligned dual-polynomial fusion, and check whether the fused 3D polynomial has a single peak at the true common AoA while also recovering each user's distinct delay and Doppler. If the fused peak falls at the average of the users' delays or at no physically realizable path, the common-3D-peak premise fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that a pilot-free multi-user OFDM uplink can perform integrated sensing and communication by solving a single 3D super-resolution problem over the unified delay–Doppler–AoA parameter space. Each user's known codebook structure lets the base station recover the matrix $X_i = \sum_\ell c_{\ell,i} f_i a_{3D}(\zeta_{\ell,i})^T$ through atomic norm minimization, implemented as a semidefinite program with a three-level Toeplitz constraint. The delay, Doppler, and AoA values are then extracted either by a Vandermonde decomposition of the Toeplitz matrix (the MaPP3D matrix-pencil and pairing procedure) or by locating peaks of the dual polynomial $f_i(\zeta)$. With multiple users observing the same target, the paper proposes fusing these dual polynomials pointwise, by maximum, by weighting, or by phase-aligning user codebooks, and reports that these collaborative estimates reduce target parameter error and symbol error rate relative to single-user or non-collaborative estimation.

Load-bearing premise

The fusion step assumes every user's measurement of a common target shares one identical 3D peak in delay-Doppler-AoA space, but physically only the angle of arrival is common; delay and Doppler depend on each user's position and velocity.

Editorial extensions

If this is right

  • If the central claim holds, uplink ISAC needs no dedicated pilots: the base station can sense targets and decode data from the same OFDM transmissions, freeing spectrum and time resources.
  • Spatially separated users provide multiple independent delay and Doppler measurements of a shared target, so target localization ambiguity shrinks as the number of users grows.
  • Communication scatterers, which cluster at near-zero Doppler, can be separated from moving targets in the 3D parameter space, enabling simultaneous communication-channel estimation and radar detection.
  • Collaborative fusion should make the system more robust to noise: the paper reports that at SNR = 0 dB, weighted, maximum, and aligned averaging all outperform single-user and non-collaborative estimation in both AoA error and aggregate symbol error rate.

Reading between the lines

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

  • A natural extension the paper leaves implicit is geometry-aware fusion: since delays and Dopplers are user-dependent in bistatic geometries, fusing in target position-velocity coordinates rather than raw parameter space could retain the collaboration benefit across users at very different locations.
  • In multi-target scenes, peak association across users becomes necessary; a naive pointwise average of dual polynomials could merge peaks from different targets. One testable extension is to cluster each user's peaks and fuse only matched targets.
  • Because the dual-polynomial SDP is solved per user or jointly, the complexity grows with user count; an online or streaming variant could make the approach practical for fast-moving users, but the paper does not address real-time implementation.
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

3 major / 5 minor

Summary. The paper proposes a pilot-free multi-user uplink ISAC framework in which single-antenna users transmit OFDM data symbols, and the base station jointly decodes the messages and estimates delay, Doppler, and angle-of-arrival parameters of sparse multipath components. The received signal is modeled as a superposition of user-specific atoms (Eq. 8), and the problem is formulated as a 3D atomic norm minimization solved by SDP, with a primal Vandermonde-decomposition route and a dual-polynomial route. The authors then propose collaborative fusion methods—pointwise averaging, pointwise maximum, weighted averaging, and codebook-aligned aggregation of the per-user dual polynomials—and claim that multi-user fusion enhances both target estimation and communication performance. Numerical experiments address continuous parameter recovery, target localization from per-user delay estimates, AoA-only collaborative estimation, and symbol error rate. The central advertised claim is that additional uplink users improve rather than degrade sensing and decoding in a unified 3D framework.

Significance. If established, the proposed framework would be a useful contribution: it turns multi-user uplink transmissions into sensing diversity and removes the need for dedicated pilots, which is attractive for mm-wave ISAC. The paper has some genuine strengths: a concrete 3D atomic norm/SDP formulation built on a structured codebook, two algorithmic routes (primal and dual), and numerical evidence that AoA-only collaboration and delay-based least-squares localization can benefit from multiple users. However, the load-bearing claim of 3D collaborative dual-polynomial fusion is not supported by the evidence and is, as stated, inconsistent with the signal model. The experimental validation also contains an oracle-dependent normalization step. The paper does not provide theoretical recovery guarantees, complexity analysis, or code, so the case for the central claim rests entirely on the simulations, which conspicuously avoid the problematic 3D fusion scenario. The work is therefore of moderate significance and requires substantial revision before its main contribution can be credited.

major comments (3)
  1. [Section IV (collaborative fusion); Eqs. (8), (32); Section V-C] The central premise of the 3D fusion methods is inconsistent with the paper's own signal model. In Eq. (8), each path has a user-dependent triple ζ_{ℓ,i}=(τ_{ℓ,i},ν_{ℓ,i},θ_{ℓ,i}), and for a common target Eq. (32) gives τ_{i,1}=(‖x_i-x_T‖+‖x_T-x_BS‖)/c, which depends on the user's location; the bistatic Doppler shift likewise depends on the user-target geometry. Therefore the dual polynomials f_i(ζ) have peaks at different (τ,ν) values for different users, and pointwise averaging, pointwise maximum, weighted averaging, and codebook-aligned aggregation on a common grid do not produce a common peak. The alignment step removes only the constant c_user(i) from the codebook, and since f_i is defined as a norm the phase of c_user(i) is lost anyway. The numerical experiments confirm that the 3D fusion is never actually demonstrated: Figure 4 restricts to AoA-only estimation with P=Q=1, and the localization experiment in Eq. (35) uses per-user delay estimates rather than fused dual polynomials. The authors should either restrict the fusion claim to genuinely common parameters (e.g., AoA or target location) and use per-user delay/Doppler estimates as inputs to a geometric fusion step, or explicitly warp/align the per-user parameter spaces before aggregation. Without such a change, the claimed 3D collaborative fusion is not validated.
  2. [Section V-B, Eq. (33)] The localization experiment normalizes the true physical delays using the ground-truth minimum and maximum over users, τ_{i,1}=(τ_{i,1}-min_i τ_{i,1})/(max_i τ_{i,1}-min_i τ_{i,1}). This is an oracle step: in a real pilot-free ISAC system these constants are not known in advance, and they depend on the unknown user and target geometry. The reported MAE reduction with increasing R in Fig. 3b therefore reflects, at least in part, the use of information that the method would not have access to. The authors should either estimate the normalization constants from the data, or study the sensitivity of the result to the choice of the affine mapping, or re-run the experiment without oracle normalization.
  3. [Section IV-B, Eq. (30)] The message-recovery step solves for the products c_{ℓ,i} f_i as independent unknowns, one per path. However, in the model (8) the same symbol vector f_i is shared by all paths of user i. As written, the least-squares problem in Eq. (30) has s_i k_i unknowns per user, whereas the model has only k_i + s_i degrees of freedom for that user. The text refers to [20, Eq. (29)] for f_i extraction, but the paper should specify how the common f_i is recovered from the s_i estimated vectors and why the least-squares solution is not adversely affected by this over-parameterization. Since the reported SER results depend directly on this step, the procedure needs to be stated precisely and validated, otherwise the communication-performance claims are not reproducible from the text alone.
minor comments (5)
  1. [Eq. (5) vs. Eq. (8)] The delay phase in Eq. (5) appears as e^{-j2πτ_{ℓ,i}q/T}, while the atom in Eq. (8) is defined as e^{+j2π(qτ+pν+rθ)} with normalized τ=τ/T. This sign inconsistency should be fixed; as written, the delay component in the atomic model is mirrored relative to the received-signal model.
  2. [Eqs. (26)-(27)] The constraint in Eq. (27) should hold for all ζ∈[0,1)^3, but the notation ζ_i∈[0,1)^3 suggests that ζ_i is a single optimization variable per user. Please correct the quantifier and the notation to avoid confusion.
  3. [Section III-B4 and Eq. (19)-(20)] The role of W_i in the SDP (19)-(20) is under-specified: it should be explicitly defined as a Hermitian matrix variable of appropriate dimension, and the text should clarify why the objective contains tr(W_i) while the constraint couples it with T3L(V_i) and Z_i.
  4. [Section III-A] The statement that the number of unknowns in (8) is O(LR+Σ_i s_i) is imprecise: the user messages contribute k_i symbols per user, and the continuous delay/Doppler/AoA parameters are not naturally counted by this expression. Please clarify the counting and the role of L and R.
  5. [General presentation] There are several typographical and formatting issues, including 'wherer' in Section II, 'paramters' in Section III-B4, the unusual notation 'Rexq,yy' in Eq. (26), and the formatting of '1l=0' in Eq. (28). These should be cleaned up.

Circularity Check

1 steps flagged · score 4.0 of 10

Ground-truth delay normalization in Section V-B makes the collaborative localization 'prediction' partly circular; the core algorithm is otherwise not circular, though the 3D fusion claim is left unvalidated.

  1. fitted input called prediction [Section V-B, Equations (32)-(35)]
    "τ_{i,1} = (||x_i − x_T||_2 + ||x_T − x_BS||_2)/c_light ... an linear mapping is applied to normalize the delays into the interval [0, 1]: τ_{i,1} = (τ_{i,1} − min_i τ_{i,1})/(max_i τ_{i,1} − min_i τ_{i,1}). This mapping ensures that the super-resolution method provided in Section III-B4 has sufficient dynamic range for delay separation."

    The normalization constants min_i τ_{i,1} and max_i τ_{i,1} are computed from Eq. (32), whose right-hand side contains the ground-truth target position x_T that the experiment claims to estimate. The estimated normalized delays are mapped back using this same ground-truth-derived affine map, so d_meas(i) ≈ ||x_i − x_T||_2 + ||x_T − x_BS||_2 before the least-squares step (35). Thus the 'predicted' d_T,BS is preconditioned by the target location through the normalization; the multi-user gain in Fig. 3b is a fitted-input-called-prediction artifact rather than a blind prediction from the signal model alone.

full rationale

Most of the derivation chain is self-contained: the signal model (8) defines y in terms of the 3D parameters and data; the atomic-norm/SDP programs (12)-(20) and the dual-polynomial construction (26)-(29) are standard convex relaxations, with the SDP equivalence cited from the published [18], an independent peer-reviewed result even though an author overlaps; message recovery via LS (30) is a straightforward inversion. No step in the central algorithm is constructed from the target output. The one genuine circularity is in the collaborative-localization experiment: Eq. (33) normalizes delays using the min and max of the true delays from Eq. (32), which already contain the target position x_T, and the estimated normalized delays are mapped back with the same ground-truth-derived affine map, so d_meas(i) is essentially the true distance-sum before the NLS step (35). This makes the Figure 3b improvement partly circular. Separately, Section V-C explicitly concedes that delay parameters must differ across users because users are at different locations, which means the pointwise 3D fusion of dual polynomials on a common (τ, ν, θ) grid is not demonstrated; this is an internal-consistency gap rather than a circular reduction, so it does not raise the circularity score further. Overall circularity score 4/10.

Assumptions & free parameters 4 free parameters · 7 assumptions · 0 invented entities

The central claim rests on sparsity assumptions and on SDP and dual-polynomial results imported from prior literature, plus an unjustified assumption that common targets have a shared 3D parameter peak across users. The localization validation additionally depends on ground-truth delay normalization, which is a free parameter set by the simulation geometry.

free parameters (4)
  • Affine delay normalization constants (min and max of true delays) = True min_i tau_i and max_i tau_i for the simulated geometry in Fig. 3b
    Eq. (33) maps physical delays into [0,1] using the true minimum and maximum over users; the estimated normalized delays are mapped back using those same ground-truth values, which are not available to a real base station.
  • Number of paths per user s_i = Known or dominant value in simulations, e.g., s1=s2=2 in Fig. 2
    MaPP3D selects s_i dominant eigenvalues; in simulations s_i is effectively the ground-truth number of paths, and the selection criterion is not specified.
  • User codebook perturbation amount for A_i = Unspecified small perturbation
    Section IV says A_i is obtained by applying a circular shift and a small perturbation to a common basis subspace; the perturbation size is not specified and may affect the alignment and fusion performance.
  • Dual polynomial grid resolution for peak finding = Not stated
    Peak positions are found on a fine 3D grid; the grid spacing is not given, and it directly controls the attainable estimation error.
assumptions (7)
  • domain assumption The mmWave channel is sparse: each user has a small number si of propagation paths, shared or user-specific.
    Used throughout Section II-B to justify atomic norm sparsity; no validation with a specific channel model or measurement is provided.
  • domain assumption Doppler shift is small relative to subcarrier spacing (nu << 1/T) so it is constant within an OFDM block.
    Invoked after Eq. (2) to approximate e^{j2πnu t} by a constant within each block; standard in OFDM but restricts target velocities.
  • standard math The 3-level Toeplitz SDP exactly represents the atomic norm for the atom set {f_i a_3D(zeta)^T}.
    Adapted from [18, Proposition 1] and multidimensional line spectral estimation [23]-[25]; no proof is given in this paper.
  • standard math The dual SDP is equivalent to the trigonometric polynomial constraint via Lagrangian duality.
    Invoked from trigonometric polynomials theory [13],[24]; no derivation is given in the text.
  • standard math The 3-level Toeplitz matrix admits a Vandermonde decomposition with si atoms.
    Relies on Vandermonde decomposition of multilevel Toeplitz matrices [25].
  • domain assumption Noise is zero-mean with variance sigma^2 and bounded norm ||epsilon||_2 <= eta.
    Used in the constraint of (13) and the choice of eta; exact noise statistics are not verified in the paper.
  • ad hoc to paper For a common target, the same 3D parameter triple (delay, Doppler, AoA) appears in all users' dual polynomials.
    Section IV assumes all users measure the same target parameters repeatedly; in bistatic geometry delay and Doppler vary per user, so this premise is questionable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of One Target, Many Views: Multi-User Fusion for Collaborative Uplink ISAC." pith.science (2026). https://pith.science/paper/6NAHONRQ

@misc{pith2026250501223,
  author       = {Pith},
  title        = {Pith review of: One Target, Many Views: Multi-User Fusion for Collaborative Uplink ISAC},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6NAHONRQ}},
  note         = {Machine review of arXiv:2505.01223}
}
read the original abstract

We propose a novel pilot-free multi-user uplink framework for integrated sensing and communication (ISAC) in mm-wave networks, where single-antenna users transmit orthogonal frequency division multiplexing signals without dedicated pilots. The base station exploits the spatial and velocity diversities of users to simultaneously decode messages and detect targets, transforming user transmissions into a powerful sensing tool. Each user's signal, structured by a known codebook, propagates through a sparse multi-path channel with shared moving targets and user-specific scatterers. Notably, common targets induce distinct delay-Doppler-angle signatures, while stationary scatterers cluster in parameter space. We formulate the joint multi-path parameter estimation and data decoding as a 3D super-resolution problem, extracting delays, Doppler shifts, and angles-of-arrival via atomic norm minimization, efficiently solved using semidefinite programming. A core innovation is multiuser fusion, where diverse user observations are collaboratively combined to enhance sensing and decoding. This approach improves robustness and integrates multi-user perspectives into a unified estimation framework, enabling high-resolution sensing and reliable communication. Numerical results show that the proposed framework significantly enhances both target estimation and communication performance, highlighting its potential for next-generation ISAC systems.

Figures

Figures reproduced from arXiv: 2505.01223 by the authors.

Figure 1
Figure 1. The triple delay-Doppler-angle recovery by [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 3
Figure 3. 3a: Multi-user collaboration to sense the common target and scatterers. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Comparison of collaborative methods with non-collaborative methods. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 17 canonical work pages

  1. [1]

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

    F. Liu, C. Masouros, A. P. Petropulu, H. Griffiths, and L. Hanzo, “Joint radar and communication design: Applications, state-of-the-art, and the road ahead,” IEEE Transactions on Communications , vol. 68, no. 6, pp. 3834–3862, 2020

    2020

  2. [2]

    Signaling strategies for dual-function radar communications: An overview,

    A. Hassanien, M. G. Amin, Y . D. Zhang, and F. Ahmad, “Signaling strategies for dual-function radar communications: An overview,” IEEE Aerospace and Electronic Systems Magazine, vol. 31, no. 10, pp. 36–45, 2016

    work page 2016

  3. [3]

    Survey on 6g frontiers: Trends, applications, require- ments, technologies and future research,

    C. D. Alwis, A. Kalla, Q.-V . Pham, P. Kumar, K. Dev, W.-J. Hwang, and M. Liyanage, “Survey on 6g frontiers: Trends, applications, require- ments, technologies and future research,” IEEE Open Journal of the Communications Society, vol. 2, pp. 836–886, 2021

    2021

  4. [4]

    Integrated sensing and communications: Toward dual-functional wire- less networks for 6g and beyond,

    F. Liu, Y . Cui, C. Masouros, J. Xu, T. X. Han, Y . C. Eldar, and S. Buzzi, “Integrated sensing and communications: Toward dual-functional wire- less networks for 6g and beyond,” IEEE journal on selected areas in communications, vol. 40, no. 6, pp. 1728–1767, 2022

    2022

  5. [5]

    A survey on fundamental limits of integrated sensing and communication,

    A. Liu, Z. Huang, M. Li, Y . Wan, W. Li, T. X. Han, C. Liu, R. Du, D. K. P. Tan, J. Lu, et al. , “A survey on fundamental limits of integrated sensing and communication,” IEEE Communications Surveys & Tutorials, vol. 24, no. 2, pp. 994–1034, 2022

    2022

  6. [6]

    Collaborative sensing in perceptive mobile networks: Opportunities and challenges,

    L. Xie, S. Song, Y . C. Eldar, and K. B. Letaief, “Collaborative sensing in perceptive mobile networks: Opportunities and challenges,” IEEE Wireless Communications, vol. 30, no. 1, pp. 16–23, 2023

    work page 2023

  7. [7]

    Energy-efficient integrated sensing and communication in collaborative millimeter wave networks,

    Y . Cui, H. Ding, Y . Ma, X. Li, H. Zhang, and Y . Fang, “Energy-efficient integrated sensing and communication in collaborative millimeter wave networks,” IEEE Transactions on Wireless Communications , vol. 24, no. 3, pp. 2341–2357, 2025

    work page 2025

  8. [8]

    Isac receiver design: Joint doa and data estimation in the presence of incomplete signal observations,

    I. Valiulahi, C. Masouros, M. Alaaeldin, and E. Alsusa, “Isac receiver design: Joint doa and data estimation in the presence of incomplete signal observations,”IEEE Open Journal of Vehicular Technology, pp. 1– 7, 2025

    work page 2025

Show all 28 references
  1. [9]

    Seventy years of radar and communications: The road from separation to integration,

    F. Liu, L. Zheng, Y . Cui, C. Masouros, A. P. Petropulu, H. Griffiths, and Y . C. Eldar, “Seventy years of radar and communications: The road from separation to integration,” IEEE Signal Processing Magazine , vol. 40, no. 5, pp. 106–121, 2023

    2023

  2. [10]

    Sensing with communication signals: From infor- mation theory to signal processing,

    F. Liu, Y .-F. Liu, Y . Cui, C. Masouros, J. Xu, T. X. Han, S. Buzzi, Y . C. Eldar, and S. Jin, “Sensing with communication signals: From infor- mation theory to signal processing,” arXiv preprint arXiv:2502.10819 , 2025

    2025 arXiv

  3. [11]

    Cooperative isac networks: Opportunities and challenges,

    K. Meng, C. Masouros, A. P. Petropulu, and L. Hanzo, “Cooperative isac networks: Opportunities and challenges,” IEEE Wireless Communi- cations, 2024

    2024

  4. [12]

    T. S. Rappaport, R. W. Heath Jr, R. C. Daniels, and J. N. Murdock, Millimeter wave wireless communications . Pearson Education, 2015

    2015

  5. [13]

    Towards a mathematical theory of super-resolution,

    E. J. Cand `es and C. Fernandez-Granda, “Towards a mathematical theory of super-resolution,” Communications on pure and applied Mathematics, vol. 67, no. 6, pp. 906–956, 2014

    2014

  6. [14]

    Compressed sensing off the grid,

    G. Tang, B. N. Bhaskar, P. Shah, and B. Recht, “Compressed sensing off the grid,” IEEE Trans. on Info. Theo. , vol. 59, no. 11, pp. 7465–7490, 2013

    2013

  7. [15]

    Super-resolution of point sources via convex programming,

    C. Fernandez-Granda, “Super-resolution of point sources via convex programming,” Information and Inference: A Journal of the IMA , vol. 5, no. 3, pp. 251–303, 2016

    2016

  8. [16]

    Two-dimensional super-resolution via convex relaxation,

    I. Valiulahi, S. Daei, F. Haddadi, and F. Parvaresh, “Two-dimensional super-resolution via convex relaxation,” IEEE Trans. on Sig. Proc. , vol. 67, no. 13, pp. 3372–3382, 2019

    2019

  9. [17]

    Blind two-dimensional super- resolution in multiple-input single-output linear systems,

    S. Sayyari, S. Daei, and F. Haddadi, “Blind two-dimensional super- resolution in multiple-input single-output linear systems,” IEEE Sig. Proc. Letters, 2020

    2020

  10. [18]

    Separating radar signals from impulsive noise using atomic norm minimization,

    S. Bayat and S. Daei, “Separating radar signals from impulsive noise using atomic norm minimization,” IEEE Transactions on Circuits and Systems II: Express Briefs , vol. 68, no. 6, pp. 2212–2216, 2020

    2020

  11. [19]

    Near-field isac in 6g: Addressing phase nonlinearity via lifted super- resolution,

    S. Daei, A. Zamani, S. Chatterjee, M. Skoglund, and G. Fodor, “Near-field isac in 6g: Addressing phase nonlinearity via lifted super- resolution,” in ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pp. 1–5, 2025

    2025

  12. [20]

    Timely and painless breakups: Off-the-grid blind message recovery and users’ demixing,

    S. Daei, S. Razavikia, M. Skoglund, G. Fodor, and C. Fischione, “Timely and painless breakups: Off-the-grid blind message recovery and users’ demixing,” arXiv preprint arXiv:2406.17393 , 2024

    2024 arXiv

  13. [21]

    Off- the-grid blind deconvolution and demixing,

    S. Razavikia, S. Daei, M. Skoglund, G. Fodor, and C. Fischione, “Off- the-grid blind deconvolution and demixing,” in GLOBECOM 2023-2023 IEEE Global Communications Conference, pp. 7604–7610, IEEE, 2023

    2023

  14. [22]

    Super-resolution delay-doppler estimation for ofdm passive radar,

    L. Zheng and X. Wang, “Super-resolution delay-doppler estimation for ofdm passive radar,” IEEE Transactions on Signal Processing , vol. 65, no. 9, pp. 2197–2210, 2017

    2017

  15. [23]

    The con- vex geometry of linear inverse problems,

    V . Chandrasekaran, B. Recht, P. A. Parrilo, and A. S. Willsky, “The con- vex geometry of linear inverse problems,”Foundations of Computational mathematics, vol. 12, no. 6, pp. 805–849, 2012

    2012

  16. [24]

    Dumitrescu, Positive Trigonometric Polynomials and Signal Process- ing Applications

    B. Dumitrescu, Positive Trigonometric Polynomials and Signal Process- ing Applications. Springer, 2017

    2017

  17. [25]

    Vandermonde decomposition of mul- tilevel toeplitz matrices with application to multidimensional super- resolution,

    Z. Yang, L. Xie, and P. Stoica, “Vandermonde decomposition of mul- tilevel toeplitz matrices with application to multidimensional super- resolution,” IEEE Transactions on Information Theory , vol. 62, no. 6, pp. 3685–3701, 2016

    2016

  18. [26]

    CVX: MATLAB software for disciplined convex programming, version 2.1,

    M. Grant and S. Boyd, “CVX: MATLAB software for disciplined convex programming, version 2.1,” 2014

    2014

  19. [27]

    Esprit-estimation of signal parameters via rota- tional invariance techniques,

    R. Roy and T. Kailath, “Esprit-estimation of signal parameters via rota- tional invariance techniques,” IEEE Transactions on acoustics, speech, and signal processing , vol. 37, no. 7, pp. 984–995, 1989

    1989

  20. [28]

    Blind goal-oriented massive access for future wireless networks,

    S. Daei and M. Kountouris, “Blind goal-oriented massive access for future wireless networks,” IEEE Transactions on Signal Processing , vol. 71, pp. 901–916, 2023

    2023

Pith tools

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