Pith. sign in

REVIEW 4 major objections 6 minor 5 cited by

A Flexible Design Framework for Integrated Communication and Computing Receivers

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper establishes a two-loop Gaussian belief propagation receiver that detects data symbols and computes AirComp functions, obtaining the combiner from a linear system solve rather than a matrix inversion.

desk verdict Solid engineering extension of GaBP to joint data/AirComp detection, but the matrix-inversion-free combiner loop lacks convergence analysis and should be validated with code and more scenarios before I'd trust it generally. read the letter →

arxiv 2506.05944 v1 pith:WYTK5C7L submitted 2025-06-06 eess.SP

Kuranage Roche Rayan Ranasinghe , Kengo Ando , Hyeon Seok Rou , Giuseppe Thadeu Freitas de Abreu , Takumi Takahashi , Marco Di Renzo , David González G This is my paper · ORCID
classification eess.SP
keywords integratedcommunicationandcomputingover-the-aircomputationGaussianbeliefpropagationmassiveMIMOnomographicfunctionmulti-streamAirCompmulti-accessreceiverlow-complexitydetect-and-combine
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 proposes a receiver design in which one base station simultaneously detects independent data symbols and computes over-the-air functions of the form $f(\mathbf{s})=\phi(\sum_k \psi_k(s_k))$, with each user able to transmit data, computing signals, or both. The receiver estimates the data vector $\mathbf{d}$ with a Gaussian belief propagation (GaBP) loop, then computes the AirComp target $\hat{f}(\mathbf{s})$ by applying a second GaBP loop to obtain the combining vector $\mathbf{u}$ from the linear system $\mathbf{b}=\mathbf{A}\mathbf{u}$, avoiding the cubic-complexity matrix inversion used by earlier combiners. The central claim is that, despite the reduced complexity $O(NK)+O(N^2)$, the method matches the benchmark's bit error rate and normalized mean square error in underloaded and fully loaded setups with up to 200 antennas and 300 users. A sympathetic reader would care because it offers a systematic, low-complexity way to integrate communication and computation at the physical layer, generalizing to arbitrary nomographic functions and multiple computation streams.

What carries the argument

The load-bearing object is the MMSE AirComp combiner $\mathbf{u}=(\mathbf{H}(\sigma_s^2\mathbf{I}_K+\boldsymbol{\Omega})\mathbf{H}^{\mathsf H}+\sigma_w^2\mathbf{I}_N)^{-1}\mathbf{H}\sigma_s^2\mathbf{1}_K$. Instead of inverting the $N\times N$ matrix, the paper recasts it as the linear system $\mathbf{b}=\mathbf{A}\mathbf{u}$ and applies Gaussian belief propagation (GaBP), a message-passing scheme for inference in Gaussian models whose soft interference cancellation and belief generation operate element-wise on scalars. Gaussian denoisers for the combiner use prior variance $\sigma_u^2\approx 1$ with mean updated by EM, while the data-detection loop uses the same GaBP machinery under a diagonal approximation of the effective noise covariance. The two loops together replace one cubic matrix inversion with $O(NK)+O(N^2)$ scalar updates, and this is what allows the scheme to scale to 200 antennas and 300 users.

What would settle it

Take the $N=100$, $K=75$ fully loaded setup at a fixed $\mathrm{SNR}_S=20$ dB and replace the i.i.d. Rayleigh channel by one with exponential correlation matrix whose off-diagonal coefficient $\rho$ is swept from 0 to 0.9; if the NMSE of $\hat{f}(\mathbf{s})$ rises significantly with $\rho$, the diagonal approximation in equation (30) is the limiting assumption.

Watch

Extended reading notes

Core claim

The paper demonstrates that the receive combiner for over-the-air computing can be treated as a linear estimation problem inside the same message-passing framework as data detection. After GaBP yields an estimate of $\mathbf{d}$, the receiver forms $\mathbf{A}=\mathbf{H}(\sigma_s^2\mathbf{I}_K+\boldsymbol{\Omega})\mathbf{H}^{\mathsf H}+\sigma_w^2\mathbf{I}_N$ and $\mathbf{b}=\mathbf{H}\sigma_s^2\mathbf{1}_K$, and obtains the MMSE combining vector by running GaBP on $\mathbf{b}=\mathbf{A}\mathbf{u}$ with a Gaussian prior on $\mathbf{u}$ and an EM update for its mean. Because the effective noise covariance $R_{\tilde{\mathbf w}}=\sigma_s^2\mathbf{H}\mathbf{H}^{\mathsf H}+\sigma_w^2\mathbf{I}_N$ is approximated as diagonal, $\sigma_s^2\mathrm{diag}(\xi)+\sigma_w^2\mathbf{I}_N$, all message updates reduce to element-wise scalar operations. The paper shows by simulation that this two-loop GaBP receiver reproduces the BER and NMSE of the benchmark joint estimator in underloaded and fully loaded settings, and extends to multiple computation streams through sparse index vectors $\mathbf{p}_m$ and to multi-access operation by zeroing the data estimates of computing-only users.

Load-bearing premise

The load-bearing premise is that the interference plus noise seen by the data detector is well approximated by independent per-antenna noise, which the paper justifies by channel hardening and small $\sigma_s^2$; when $\mathbf{H}\mathbf{H}^{\mathsf H}$ has strong off-diagonal entries, the soft interference-cancellation variances are miscalibrated and both data detection and the computed function degrade.

Editorial extensions

If this is right

  • Data detection and over-the-air computing can share one receiver with no cubic matrix inversion: the communication loop costs $O(NK)$ and the combining loop costs $O(N^2)$.
  • The combiner design is agnostic to the target function, so any nomographic function supported by suitable pre- and post-processing can be computed with the same receiver structure.
  • Multiple target functions can be computed from the same set of computing symbols by defining sparse index vectors $\mathbf{p}_m$ and running the combining loop once per stream.
  • Multi-access operation requires only zeroing data estimates for computing-only users, so communication-only, computing-only, and hybrid users coexist with minimal structural change.
  • In overloaded regimes, AirComp NMSE degrades, and the paper attributes the degradation to the combiner rather than to data detection.

Reading between the lines

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

  • Because the diagonal effective-noise approximation is what keeps the updates scalar, a natural test is to sweep the spatial correlation of the channel: if NMSE rises sharply as $\mathbf{H}\mathbf{H}^{\mathsf H}$ gains off-diagonal weight, that confirms the approximation as the bottleneck.
  • The combining loop solves a Hermitian positive-definite linear system, so preconditioning or low-rank approximations of $\mathbf{A}$ could plausibly push the cost below $O(N^2)$ without changing the message-passing structure.
  • The paper states that correlated computation streams would require dedicated conditional-probability message passing; a decorrelating preprocessing step might preserve scalar updates and extend the framework to correlated streams at modest cost.
  • The claim that overloaded AirComp is combiner-limited could be checked by replacing the MMSE combiner with a regularized variant and measuring whether the NMSE floor in the $K>N$ regime drops.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript proposes a Gaussian belief propagation (GaBP) framework for integrated communication and computing (ICC) receivers in a SIMO uplink. The receiver jointly estimates the data vector d and evaluates a nomographic function f(s)=φ(Σ_k ψ_k(s_k)) over-the-air. A benchmark algorithm first detects both d and the per-user computing signals s, then applies an MMSE combiner that requires an N×N matrix inversion. The proposed single-stream method treats Hs+w as an approximately diagonal effective noise, detects d by GaBP with soft interference cancellation, and replaces the combiner matrix inversion with a second GaBP loop that solves b=Au. The framework is then extended to multiple independent computing streams and to a multi-access setup where users transmit data, computing, or both. Simulations for systems with up to 200 antennas and 300 users show that the proposed algorithms match the benchmark BER/NMSE and approach matched-filter bounds in underloaded and fully loaded cases.

Significance. Assuming the two central approximations hold, the framework is a valuable step toward scalable ICC receivers: the MMSE combiner derivation in Appendix VII-A is explicit and self-contained, the GaBP updates are element-wise, and the multi-stream/multi-access extension is clean and general. The simulations cover a wider range of loading conditions than many AirComp papers and demonstrate that the proposed method approaches the matched-filter bound in the under/fully loaded cases. The main reservations are that the diagonal effective-noise approximation and the GaBP solution of the noiseless linear system are not supported by any formal analysis, and the prior variance and EM update in the combining loop appear to be mismatched to the problem; these points need to be addressed before the generality and complexity claims can be accepted.

major comments (4)
  1. [Section IV-A, Eq. (30)] The approximation R_tilde_w ≈ σ_s^2 diag(ξ) + σ_w^2 I is load-bearing for the soft-IC variances in Eq. (35), but its stated justification is incomplete. For i.i.d. CN(0,1) channels and E_S = E_D/K, the off-diagonal entries of σ_s^2 H H^H have zero mean and standard deviation σ_s^2 sqrt(K) = E_D/sqrt(K), while the diagonal entries have mean σ_s^2 K = E_D. Hence at high SNR the relative off-diagonal magnitude is 1/sqrt(K), which is 0.11 for K=75 and 0.07 for K=200; the approximation is an uncontrolled O(1/sqrt(K)) neglect of correlated interference. The paper should either bound the resulting error in the variances of Eq. (35) or explicitly restrict the claim to regimes where this error is negligible.
  2. [Section IV-B, Eqs. (46)-(51), Algorithm 2 steps 7-12] The claim that the combining vector can be obtained by a second GaBP loop over b = A u lacks a convergence guarantee. Unlike the data-detection loop, this system has no observation noise, so the likelihood variance in Eq. (51) is purely residual interference left by imperfect estimates; as the estimates improve, these variances shrink, and the messages become overconfident. The standard GaBP convergence conditions of reference [48] are not checked for the dense matrix A, and the paper gives no argument that the loop's fixed point equals A^{-1} b. Without such support, the NMSE agreement in Fig. 3 is evidence only for the simulated parameter sets, not for the general equivalence of the GaBP combiner to the exact MMSE combiner.
  3. [Section IV-B, Eqs. (55)-(58)] The Gaussian prior used in the combining loop is initialized with σ_u^2 = 1 and μ_u = 0, but for i.i.d. Gaussian H and C ≈ σ_s^2 I, the entries of u = A^{-1} b have typical per-entry variance of order K/[N(N-K)] for N > K (about 0.03 for N=100, K=75), not 1. The prior in Eq. (55) therefore biases the solution, and the EM update in Eq. (58), which averages all N^2 extrinsic estimates into a single mean μ_u, has no derivation: the entries of u are distinct random variables, not i.i.d. samples of a common parameter. This step should be either derived, replaced by a non-informative initialization, or reported as a tuned heuristic.
  4. [Section IV-C, Algorithm 2 combining stage] The claimed O(N^2) complexity for the combining stage is not supported by the equations as written. For each of the N^2 pairs (n,n'), Eqs. (49) and (51) require a sum over q≠n' of N terms, and Eqs. (53) and (54) require a sum over q≠n of N terms; a direct implementation is therefore O(N^3) per iteration. The O(N^2) complexity is achievable only by maintaining row-wise totals such as Σ_q a_{n,q} \hat u^{(i-1)}_{n,q} and subtracting the self term for each message; this schedule should be stated explicitly, together with the number of combining iterations, before the complexity claim can be assessed.
minor comments (6)
  1. [Section IV-B, before Eq. (47)] The sentence refers to 'equation (65)', which does not exist; it should refer to Eq. (46).
  2. [Algorithm 2, initialization] The combiner-loop variables \hat u^{(0)}_{n,n'} and \hat σ^{2(0)}_{u:n,n'} are not initialized, so the pseudocode is not fully reproducible.
  3. [Algorithm 2, step 6] Step 6 says to compute A and b from Eq. (44), but the data-error covariance Ω entering A is not defined in that section; it should be specified (as in Section III-B) in terms of the data-detection variances.
  4. [Section IV-C, simulation parameters] The damping factor β_u is listed as an input to Algorithms 2 and 3, but its value is not reported in the simulations, so the numerical results cannot be reproduced without code.
  5. [Footnote 9 and Figs. 2, 3, 6] Footnote 9 notes that joint estimation of d and s is overloaded even for K<N, which conflicts with the 'underloaded' label used in the figures and text; the terminology should be reconciled.
  6. [Abstract and simulation scope] The abstract's claim of 'up to 200 antennas and 300 UEs/EDs' is not supported by the proposed-method simulations: Fig. 3 reaches N=200 with K=200, and Fig. 4 reaches K=120; the 300-user point appears only in the benchmark results of Fig. 2.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the receiver derivation is self-contained and its claims do not reduce to fitted inputs or self-citation chains.

full rationale

The paper's central derivation chain is self-contained. The MMSE combiner is derived in the appendix from the stated signal model (eqs. 68-74), the data-detection GaBP rules follow from the linear model with the effective noise approximation, and the combining GaBP loop is presented as a numerical solver for the linear system b = A u (eq. 46), not as a prediction fitted to a target. Algorithm constants such as beta_d, beta_s, i_max, and sigma_s^2 = E_D/K are fixed in advance; no parameter is tuned against the benchmark BER/NMSE curves. The only self-citation to the authors' prior work [1] is as a benchmark, and Algorithm 1 is fully specified in the paper itself, so matching it is a simulation comparison rather than a load-bearing appeal to an unverified external result. The diagonal covariance approximation in eq. (30) and the unverified convergence of the noiseless GaBP loop are correctness risks and potential limitations, but they are not circular: the equations do not reduce to their own inputs, and no fitted quantity is renamed as a prediction. Accordingly, the circularity score is 0.

Assumptions & free parameters 5 free parameters · 8 assumptions · 0 invented entities

The central claim depends on a small number of hand-chosen algorithm parameters (damping factors, iteration count, sigma_u^2) and on standard domain assumptions (perfect CSI, synchronization, Gaussian computing signals, channel hardening). None of the parameters are fitted to data to force agreement; they are convergence or modeling choices. The most fragile item is the ad hoc sigma_u^2 = 1 prior for the combiner GaBP.

free parameters (5)
  • Damping factor beta_d = 0.5
    Hand-tuned to stabilize GaBP convergence; reported for benchmark simulations in Section III-D, used in all algorithms.
  • Damping factor beta_s = 0.8
    Hand-tuned for computing-signal denoiser in Algorithm 1; Section III-D reports only for the benchmark.
  • Damping factor beta_u = not reported
    Used in Algorithms 2 and 3 for the combiner GaBP update, but its value is not given in the simulation parameters, hindering reproduction.
  • Combiner prior variance sigma_u^2 = 1
    Set by hand in Section IV-B as 'usually constrained' rather than derived from the moments of u = A^{-1} b; a key assumption for the GaBP combiner denoiser.
  • Maximum iterations i_max = 30
    Stopping criterion for the GaBP loops; Section III-D states i_max = 30. Performance may depend on this choice.
assumptions (8)
  • domain assumption Channel hardening: H H^H is approximated by its diagonal diag(xi) in the effective noise covariance.
    Used in equation (30) to justify R_w_tilde approximately sigma_s^2 diag(xi) + sigma_w^2 I. Requires N large and sigma_s^2 small; not guaranteed for moderate N or correlated channels.
  • domain assumption Perfect channel state information and noise variance at the receiver.
    The receiver knows H and sigma_w^2; no estimation error or pilot overhead is modeled.
  • domain assumption Perfect symbol synchronization among all user transmissions.
    Equation (1) and the whole model rely on synchronized arrivals; asynchronous transmissions would break the linear superposition.
  • domain assumption Scalar Gaussian approximation (SGA) of interference plus noise.
    Used in equations (11), (34), and (50) to turn interference terms into Gaussian noise; standard but approximate.
  • domain assumption Computing signals s_k are complex Gaussian with known variance sigma_s^2 and unknown mean estimated by EM.
    Assumed in Section III; basis for denoiser (17a) and EM update (21).
  • domain assumption Data symbols are QPSK with known power E_D.
    The Bayes-optimal denoiser (16) is QPSK-specific; other constellations require separate denoisers as noted in footnote 7.
  • domain assumption Multi-stream functions are uncorrelated and each user belongs to exactly one stream.
    Section V-A uses disjoint support vectors p_m; relaxation is stated to be out of scope.
  • ad hoc to paper Combiner vector u has Gaussian prior with sigma_u^2 = 1.
    Section IV-B states u's variance is 'usually constrained' and sets it to 1; not derived from A and b, so a mismatched prior could bias the GaBP solution.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Flexible Design Framework for Integrated Communication and Computing Receivers." pith.science (2026). https://pith.science/paper/WYTK5C7L

@misc{pith2026250605944,
  author       = {Pith},
  title        = {Pith review of: A Flexible Design Framework for Integrated Communication and Computing Receivers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WYTK5C7L}},
  note         = {Machine review of arXiv:2506.05944}
}
read the original abstract

We propose a framework to design integrated communication and computing (ICC) receivers capable of simultaneously detecting data symbols and performing over-the-air computing (AirComp) in a manner that: a) is systematically generalizable to any nomographic function, b) scales to a massive number of user equipments (UEs) and edge devices (EDs), c) supports the computation of multiple independent functions (streams), and d) operates in a multi-access fashion whereby each transmitter can choose to transmit either data symbols, computing signals or both. For the sake of illustration, we design the proposed multi-stream and multi-access method under an uplink setting, where multiple single-antenna UEs/EDs simultaneously transmit data and computing signals to a single multiple-antenna base station (BS)/access point (AP). Under the communication functionality, the receiver aims to detect all independent communication symbols while treating the computing streams as aggregate interference which it seeks to mitigate; and conversely, under the computing functionality, to minimize the distortion over the computing streams while minimizing their mutual interference as well as the interference due to data symbols. To that end, the design leverages the Gaussian belief propagation (GaBP) framework relying only on element-wise scalar operations coupled with closed-form combiners purpose-built for the AirComp operation, which allows for its use in massive settings, as demonstrated by simulation results incorporating up to 200 antennas and 300 UEs/EDs. The efficacy of the proposed method under different loading conditions is also evaluated, with the performance of the scheme shown to approach fundamental limiting bounds in the under/fully loaded cases.

Figures

Figures reproduced from arXiv: 2506.05944 by the authors.

Figure 1
Figure 1. SIMO ICC system consisting of one BS/AP with [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. BER and NMSE achieved by the benchmark scheme (Algorithm 1) [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. BER and NMSE achieved by Algorithm 2 in underloaded and fully [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: NMSE achieved by Algorithms 1 and 2, as a function of the user [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: SIMO ICC system computing a vector of two distinct functions [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 7
Figure 7. Figure 7: SIMO ICC system computing a vector of distinct functions from the [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: BER and NMSE performance of the proposed single-stream Algorithm [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Affine Filter Bank Modulation (AFBM): A Novel 6G ISAC Waveform with Low PAPR and OOBE

    eess.SP 2025-09 conditional novelty 5.0 of 10

    AFBM, a filter-banked chirp-precoded waveform, achieves lower PAPR and OOBE than AFDM in simulation while supporting a belief-propagation data receiver and an EM-assisted sensing receiver.

  2. Doubly-Dispersive Continuous MIMO Systems: Channel Modeling and Beamforming Design

    eess.SP 2025-08 reject novelty 5.0 of 10

    A doubly-dispersive continuous MIMO channel model with matched-filter-like beamforming is derived, but the variational optimality proof is invalid because the unconstrained problem is unbounded.

  3. Low-Complexity Receiver Design for Affine Filter Bank Modulation

    eess.SP 2025-06 conditional novelty 5.0 of 10

    A Gaussian Belief Propagation detector for AFBM is shown, via simulation, to outperform AFDM by about 2 dB at BER 10^-3 in doubly-dispersive channels at low per-iteration cost.

  4. Integrated Communication and Computing in Time-Varying mmWave Channels

    eess.SP 2025-09 conditional novelty 4.0 of 10

    A BiGaBP-based receiver jointly tracks channel, detects data, and computes an over-the-air sum in time-varying mmWave channels, with simulations showing small gaps to genie-aided bounds.

  5. Flexible Intelligent Metasurfaces in High-Mobility MIMO Integrated Sensing and Communications

    eess.SP 2025-07 conditional novelty 4.0 of 10

    A flexible-intelligent-metasurface-parameterized doubly dispersive MIMO channel model is proposed, and optimizing the surface shape at both link ends is shown by simulation to improve achievable rate and angle-of-arri...

Reference graph

Works this paper leans on

50 extracted references · 48 canonical work pages · cited by 5 Pith papers

  1. [48]

    On Convergence Conditions of Gaussian Belief Propaga- tion,

    Q. Suet al., “On Convergence Conditions of Gaussian Belief Propaga- tion,”IEEE Transactions on Signal Processing, vol. 63, no. 5, 2015

    work page 2015

  2. [1]

    From theory to reality: A design framework for integrated communication and computing receivers,

    K. R. R. Ranasinghe, K. Ando, and G. T. Freitas de Abreu, “From theory to reality: A design framework for integrated communication and computing receivers,” in2025 International Conference on Computing, Networking and Communications (ICNC), 2025, pp. 865–870

    work page 2025

  3. [2]

    On the Road to 6G: Visions, Requirements, Key Technologies, and Testbeds,

    C.-X. Wanget al., “On the Road to 6G: Visions, Requirements, Key Technologies, and Testbeds,”IEEE Communications Surveys & Tutorials, vol. 25, no. 2, 2023

    work page 2023

  4. [3]

    Key Focus Areas and Enabling Technologies for 6G,

    C. G. Brintonet al., “Key Focus Areas and Enabling Technologies for 6G,”IEEE Communications Magazine, vol. 63, no. 3, 2025

    work page 2025

  5. [4]

    Six Integration Avenues for ISAC in 6G and Beyond,

    N. Gonz ´alez-Prelcicet al., “Six Integration Avenues for ISAC in 6G and Beyond,”IEEE V ehicular Technology Magazine, vol. 20, no. 1, 2025

    work page 2025

  6. [5]

    Near-real-time 6G service operation enabled by distributed intelligence and in-band telemetry,

    P. Gonzalezet al., “Near-real-time 6G service operation enabled by distributed intelligence and in-band telemetry,”Journal of Optical Com- munications and Networking, vol. 17, no. 3, 2025

    work page 2025

  7. [6]

    Drone-Enabled Connectivity: Advancements and Chal- lenges in B5G / 6G Networks,

    T. Khaledet al., “Drone-Enabled Connectivity: Advancements and Chal- lenges in B5G / 6G Networks,” inProc. 8th International Conference on Image and Signal Processing and their Applications (ISPA), 2024

    work page 2024

  8. [7]

    Toward 6G-Enabled URLLCs: Digital Twin, Open Ran, and Semantic Communications,

    A. Masaracchiaet al., “Toward 6G-Enabled URLLCs: Digital Twin, Open Ran, and Semantic Communications,”IEEE Communications Standards Magazine, vol. 9, no. 1, 2025

    work page 2025

Show all 50 references
  1. [8]

    Internet of Things (IoT) for Next-Generation Smart Systems: A Review of Current Challenges, Future Trends and Prospects for Emerging 5G-IoT Scenarios,

    K. Shafiqueet al., “Internet of Things (IoT) for Next-Generation Smart Systems: A Review of Current Challenges, Future Trends and Prospects for Emerging 5G-IoT Scenarios,”IEEE Access, vol. 8, 2020. 13

    2020

  2. [9]

    On the Effectiveness of OTFS for Joint Radar Param- eter Estimation and Communication,

    L. Gaudioet al., “On the Effectiveness of OTFS for Joint Radar Param- eter Estimation and Communication,”IEEE Transactions on Wireless Communications, vol. 19, no. 9, 2020

    2020

  3. [10]

    Joint Design of Communication and Sensing for Beyond 5G and 6G Systems,

    T. Wild, V . Braun, and H. Viswanathan, “Joint Design of Communication and Sensing for Beyond 5G and 6G Systems,”IEEE Access, vol. 9, 2021

    2021

  4. [11]

    Integrated Sensing and Communications: Toward Dual- Functional Wireless Networks for 6G and Beyond,

    F. Liuet al., “Integrated Sensing and Communications: Toward Dual- Functional Wireless Networks for 6G and Beyond,”IEEE Journal on Selected Areas in Communications, vol. 40, no. 6, 2022

    2022

  5. [12]

    Asymmetric Bilinear Inference for Joint Communica- tions and Environment Sensing,

    H. S. Rouet al., “Asymmetric Bilinear Inference for Joint Communica- tions and Environment Sensing,” inProc. 56th Asilomar Conference on Signals, Systems, and Computers, 2022

    2022

  6. [13]

    Efficient Joint Radar and Communication Exploiting Sparsity and Spatial Modulation,

    A. E. Maiet al., “Efficient Joint Radar and Communication Exploiting Sparsity and Spatial Modulation,” inProc. IEEE 9th International Work- shop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), 2023

    2023

  7. [14]

    Robust Received Signal Strength Indicator (RSSI)- Based Multitarget Localization via Gaussian Process Regression,

    N. F ¨uhrlinget al., “Robust Received Signal Strength Indicator (RSSI)- Based Multitarget Localization via Gaussian Process Regression,”IEEE Journal of Indoor and Seamless Positioning and Navigation, vol. 1, 2023

    2023

  8. [15]

    Fast and Efficient Sequential Radar Parameter Estimation in MIMO-OTFS Systems,

    K. R. R. Ranasinghe, H. S. Rou, and G. T. F. de Abreu, “Fast and Efficient Sequential Radar Parameter Estimation in MIMO-OTFS Systems,” inProc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2024

    2024

  9. [16]

    Integrated Sensing and Communications with Affine Frequency Division Multiplexing,

    A. Bemani, N. Ksairi, and M. Kountouris, “Integrated Sensing and Communications with Affine Frequency Division Multiplexing,”IEEE Wireless Communications Letters, 2024

    2024

  10. [17]

    Joint Channel, Data, and Radar Parameter Estimation for AFDM Systems in Doubly-Dispersive Channels,

    K. R. R. Ranasingheet al., “Joint Channel, Data, and Radar Parameter Estimation for AFDM Systems in Doubly-Dispersive Channels,”IEEE Transactions on Wireless Communications, vol. 24, no. 2, 2025

    2025

  11. [18]

    The Integrated Sensing and Communication Revolution for 6G: Vision, Techniques, and Applications,

    N. Gonz ´alez-Prelcicet al., “The Integrated Sensing and Communication Revolution for 6G: Vision, Techniques, and Applications,”Proceedings of the IEEE, 2024

    2024

  12. [19]

    Blind bistatic radar parameter estimation in doubly- dispersive channels,

    K. R. R. Ranasinghe, K. Ando, H. S. Rou, G. T. F. de Abreu, and A. Bathelt, “Blind bistatic radar parameter estimation in doubly- dispersive channels,” in2025 IEEE Wireless Communications and Networking Conference (WCNC), 2025, pp. 1–6

    2025

  13. [20]

    H. S. Rouet al., “From Orthogonal Time-Frequency Space to Affine Frequency-Division Multiplexing: A comparative study of next- generation waveforms for integrated sensing and communications in doubly dispersive channels,”IEEE Signal Processing Magazine, vol. 41, no. 5, 2024

    2024

  14. [21]

    Integrated Sensing and Communications for 3D Object Imaging via Bilinear Inference,

    ——, “Integrated Sensing and Communications for 3D Object Imaging via Bilinear Inference,”IEEE Transactions on Wireless Communications, vol. 23, no. 8, 2024

    2024

  15. [22]

    Intelligent Waveform Design for Integrated Sensing and Communication,

    J. Zhanget al., “Intelligent Waveform Design for Integrated Sensing and Communication,”IEEE Wireless Communications, 2024

    2024

  16. [23]

    Tone reservation-based papr reduction using manifold optimization for ofdm-isac systems,

    G. Rexhepiet al., “Tone reservation-based papr reduction using manifold optimization for ofdm-isac systems,” in2025 IEEE Wireless Communi- cations and Networking Conference (WCNC), 2025, pp. 1–6

    2025

  17. [24]

    Computation Over Multiple-Access Chan- nels,

    B. Nazer and M. Gastpar, “Computation Over Multiple-Access Chan- nels,”IEEE Transactions on Information Theory, vol. 53, no. 10, 2007

    2007

  18. [25]

    Over-the-Air Computation for 6G: Foundations, Tech- nologies, and Applications,

    Z. Wanget al., “Over-the-Air Computation for 6G: Foundations, Tech- nologies, and Applications,”IEEE Internet of Things Journal, vol. 11, no. 14, Jul. 2024

    2024

  19. [26]

    Integrated Sensing, Communication, and Computation for Over-the-Air Federated Learning in 6G Wireless Networks,

    M. Duet al., “Integrated Sensing, Communication, and Computation for Over-the-Air Federated Learning in 6G Wireless Networks,”IEEE Internet of Things Journal, vol. 11, no. 21, 2024

    2024

  20. [27]

    Cost Minimization of Integrated Sensing, Communication, and Computing in UA V-Enabled Systems,

    Z. Wang, H. Meng, Y . Cao, D. Cui, and Z. Chang, “Cost Minimization of Integrated Sensing, Communication, and Computing in UA V-Enabled Systems,” inProc. 2nd International Conference On Mobile Internet, Cloud Computing and Information Security (MICCIS), 2024

    2024

  21. [28]

    Beamforming Design for Integrated Sensing, Over-the- Air Computation, and Communication in Internet of Robotic Things,

    K. Donget al., “Beamforming Design for Integrated Sensing, Over-the- Air Computation, and Communication in Internet of Robotic Things,” IEEE Internet of Things Journal, vol. 11, no. 20, 2024

    2024

  22. [29]

    A Survey on Integrated Sensing, Communication, and Computation,

    D. Wenet al., “A Survey on Integrated Sensing, Communication, and Computation,”IEEE Communications Surveys & Tutorials, 2024

    2024

  23. [30]

    MIMO Over-the-Air Computation for High- Mobility Multimodal Sensing,

    G. Zhu and K. Huang, “MIMO Over-the-Air Computation for High- Mobility Multimodal Sensing,”IEEE Internet of Things Journal, vol. 6, no. 4, 2019

    2019

  24. [31]

    Over-the-Air Computation Systems: Optimization, Analysis and Scaling Laws,

    W. Liu, X. Zang, Y . Li, and B. Vucetic, “Over-the-Air Computation Systems: Optimization, Analysis and Scaling Laws,”IEEE Transations on Wireless Communications, vol. 19, no. 8, 2020

    2020

  25. [32]

    Over-the-Air Computation via Broadband Channels,

    T. Qinet al., “Over-the-Air Computation via Broadband Channels,” IEEE Wireless Communications Letters, vol. 10, no. 10, 2021

    2021

  26. [33]

    Simultaneous Signal-and- Interference Alignment for Two-Cell Over-the-Air Computation,

    Q. Lan, H. S. Kang, and K. Huang, “Simultaneous Signal-and- Interference Alignment for Two-Cell Over-the-Air Computation,”IEEE Communications Letters, vol. 9, no. 9, Sep. 2020

    2020

  27. [34]

    A Uniform-Forcing Transceiver Design for Over-the-Air Function Computation,

    L. Chen, X. Qin, and G. Wei, “A Uniform-Forcing Transceiver Design for Over-the-Air Function Computation,”IEEE Wireless Communica- tions Letters, vol. 7, no. 6, 2018

    2018

  28. [35]

    Optimal Receive Beamforming for Over-the-Air Computation,

    W. Fanget al., “Optimal Receive Beamforming for Over-the-Air Computation,” inProc. IEEE 22nd International Workshop on Signal Processing Advances in Wireless Communications (SPA WC), 2021

    2021

  29. [36]

    Low-Complexity and High- Performance Combiners for Over the Air Computing,

    K. Ando and G. T. F. de Abreu, “Low-Complexity and High- Performance Combiners for Over the Air Computing,” inProc. IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), 2023

    2023

  30. [37]

    Intelligent Reflecting Surface Aided AirComp: Multi- Timescale Design and Performance Analysis,

    G. Chenet al., “Intelligent Reflecting Surface Aided AirComp: Multi- Timescale Design and Performance Analysis,”IEEE Transactions on V ehicular Technology, 2024

    2024

  31. [38]

    Over-the-Air Computation via 2-D Movable Antenna Array,

    N. Liet al., “Over-the-Air Computation via 2-D Movable Antenna Array,”IEEE Wireless Communications Letters, vol. 14, no. 1, 2025

    2025

  32. [39]

    Digital Over-the-Air Computation: Achieving High Reliability via Bit-Slicing,

    J. Liu, Y . Gong, and K. Huang, “Digital Over-the-Air Computation: Achieving High Reliability via Bit-Slicing,”IEEE Transactions on Wireless Communications, 2025

    2025

  33. [40]

    Low-Complexity OTFS- Based Over-the-Air Computation Design for Time-Varying Channels,

    X. Huang, H. Hellstr ¨om, and C. Fischione, “Low-Complexity OTFS- Based Over-the-Air Computation Design for Time-Varying Channels,” IEEE Transactions on Wireless Communications, 2025

    2025

  34. [41]

    Integrated Communication and Computation Empowered by Fluid Antenna Array,

    S. Yeet al., “Integrated Communication and Computation Empowered by Fluid Antenna Array,” inProc. IEEE 25th International Workshop on Signal Processing Advances in Wireless Communications (SPA WC), 2024

    2024

  35. [42]

    Integrated Sensing, Computation, and Communication: System Framework and Performance Optimization,

    Y . He, G. Yu, Y . Cai, and H. Luo, “Integrated Sensing, Computation, and Communication: System Framework and Performance Optimization,” IEEE Transactions on Wireless Communications, vol. 23, no. 2, 2024

    2024

  36. [43]

    Gaussian Belief Propagation: Theory and Aplication,

    D. Bickson, “Gaussian Belief Propagation: Theory and Aplication,” 2009

    2009

  37. [44]

    Distributed Inference With Variational Message Passing in Gaussian Graphical Models: Tradeoffs in Message Schedules and Convergence Conditions,

    B. Li, N. Wu, and Y .-C. Wu, “Distributed Inference With Variational Message Passing in Gaussian Graphical Models: Tradeoffs in Message Schedules and Convergence Conditions,”IEEE Transactions on Signal Processing, vol. 72, 2024

    2024

  38. [45]

    Design of Adaptively Scaled Belief in Multi-Dimensional Signal Detection for Higher-Order Modulation,

    T. Takahashi, S. Ibi, and S. Sampei, “Design of Adaptively Scaled Belief in Multi-Dimensional Signal Detection for Higher-Order Modulation,” IEEE Transactions on Communications, vol. 67, no. 3, 2019

    2019

  39. [46]

    Waveforms for Computing Over the Air,

    A. P ´erez-Neira, M. Martinez-Gost, A. S ¸ahin, S. Razavikia, C. Fischione, and K. Huang, “Waveforms for Computing Over the Air,” 2024. [Online]. Available: https://arxiv.org/abs/2405.17007

    2024

  40. [47]

    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

  41. [49]

    Aoa estimation- aided bayesian receiver design via bilinear inference for mmwave massive mimo,

    K. Ito, T. Takahashi, K. Igarashi, S. Ibi, and S. Sampei, “Aoa estimation- aided bayesian receiver design via bilinear inference for mmwave massive mimo,” inICC 2023 - IEEE International Conference on Communications, 2023, pp. 6474–6479

    2023

  42. [50]

    Expectation-maximization gaussian-mixture approximate message passing,

    J. P. Vila and P. Schniter, “Expectation-maximization gaussian-mixture approximate message passing,”IEEE Transactions on Signal Processing, vol. 61, no. 19, pp. 4658–4672, 2013

    2013

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.