REVIEW 4 major objections 3 minor
AMP-based Joint Activity Detection and Channel Estimation for Massive Grant-Free Access in OFDM-based Wideband Systems
T0 review · 4 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read The paper claims that two approximate message passing algorithms, AMP-A-EC and AMP-A-AC, jointly approximate the optimal activity detection and channel estimation for OFDM-based grant-free wideband access under frequency-selective fading.
desk verdict A plausible AMP-based contribution to OFDM grant-free access, but the abstract alone leaves all load-bearing claims unverifiable; worth a full-text referee. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central object is a factor graph that exactly captures the statistics of the time-domain channels and device activities. This factor graph encodes the joint posterior distribution used in the MAP activity detection and MMSE channel estimation objectives. The approximate message passing algorithms iterate on this graph to approximate the desired estimates; the state evolution analysis of AMP-A-EC is the analytic tool that predicts its detection error probability and channel estimation MSE.
What would settle it
Simulate grant-free access with channel taps drawn from a distribution that differs from the prior assumed in the factor graph (for example, correlated taps or non-Gaussian fading), then compare AMP-A-EC's empirical detection error probability and MSE with its state-evolution prediction. A systematic gap that grows with the mismatch would show the state-evolution analysis and the algorithm's optimality claim depend on the exact prior.
Extended reading notes
Core claim
The paper's claim is that two AMP-based algorithms, AMP-A-EC and AMP-A-AC, approximately solve the MAP-based device activity detection problem and two MMSE-based channel estimation problems for OFDM-based grant-free wideband access under frequency-selective fading. AMP-A-EC is analyzed through state evolution, yielding predicted error probability and mean square error, while AMP-A-AC is shown to have lower computational complexity in the dominant term. Both algorithms fix AMP's known convergence problem when pilot length is small or comparable to the number of active devices. The paper further claims numerical evidence that both algorithms outperform existing methods in accuracy or computation time, with different preferable regions for each.
Load-bearing premise
The methods rely on the factor graph correctly matching the true statistics of time-domain channels and device activities; if the real channel or activity distribution differs from the modeled one, the derived MAP and MMSE objectives are mis-specified and the claimed performance may degrade.
Editorial extensions
If this is right
- If the claims hold, OFDM-based grant-free wideband systems can jointly detect active devices and estimate frequency-selective channels without first scheduling users, using algorithms that are both accurate and fast enough for short pilot lengths.
- The state-evolution analysis for AMP-A-EC gives a closed-form prediction of detection error probability and channel estimation MSE, so system designers can tune pilot length and thresholds analytically rather than by simulation.
- The lower dominant-term complexity of AMP-A-AC makes it the better choice in regimes where computation time is the bottleneck, while AMP-A-EC is preferable where accuracy matters.
- Both algorithms extend approximate message passing to the exact time-domain model of OFDM grant-free access, removing the convergence failure that standard AMP suffers when the pilot length is comparable to or smaller than the number of active devices.
Reading between the lines
- A natural extension the authors do not explore is to adapt these algorithms to time-varying channels or to multi-antenna base stations; the factor-graph formulation would need an added dimension for antennas or Doppler, but the same message-passing structure could carry over.
- The state-evolution result suggests a practical test: one could use AMP-A-EC's predicted error probability to compute the smallest pilot length that meets a target detection reliability, a quantity not explicitly derived in the paper.
- Because both algorithms are derived from a factor graph that assumes a particular prior, their performance under real measured channel statistics is an empirical question the paper does not settle; applying them to field data would be a direct test of the model-mismatch limit.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper addresses joint device activity detection and channel estimation for OFDM-based grant-free access in wideband systems under frequency-selective fading. It proposes an exact time-domain signal model, formulates a MAP-based activity detection problem and two MMSE-based channel estimation problems, constructs a factor graph capturing the true statistics, and derives two AMP-based algorithms, AMP-A-EC and AMP-A-AC. The abstract claims that AMP-A-EC is analyzed via state evolution, that AMP-A-AC has lower dominant-term computational complexity, and that numerical results show superior performance with distinct preferable regions for the two algorithms.
Significance. If the claimed results are fully substantiated, the work would be practically relevant for massive grant-free access in wideband OFDM systems, especially the exact time-domain model and the new MAP/MMSE formulations. The paper does not appear to rely on fitted parameters; the analytical claims rest on standard model assumptions. However, the available material is only the abstract, so the significance cannot be confirmed without examining the full derivations, state-evolution equations, and numerical evidence.
major comments (4)
- [Abstract] The central claim that AMP-A-EC and AMP-A-AC approximately solve the stated MAP/MMSE problems is unsupported in the provided text. No derivations, update equations, or convergence conditions are shown. The authors should provide the factor-graph construction, the AMP update equations, and a precise statement of the conditions under which these algorithms converge to the MAP/MMSE solutions.
- [Abstract] The claimed state-evolution analysis for AMP-A-EC is asserted but not presented. The authors should give the state-evolution recursion, state its assumptions (e.g., large-system limit, pilot statistics), and demonstrate that its predictions match finite-dimensional simulations for the OFDM wideband model.
- [Abstract] The complexity claim that AMP-A-AC has 'lower computational complexity (in dominant term)' lacks a baseline and a derivation. The authors should specify the reference algorithm, provide per-iteration complexity expressions, and state how the dominant term is obtained.
- [Abstract] The numerical results are summarized only as 'superior performance and respective preferable regions.' The authors should report the simulation setup, baseline methods, performance metrics (e.g., activity detection error probability, channel estimation MSE), and define the parameter regions in which each algorithm is preferred.
minor comments (3)
- [Abstract] The acronyms EC and AC in the algorithm names AMP-A-EC and AMP-A-AC are not defined.
- [Abstract] The phrase 'preferable regions' is vague; specifying whether these are regions in the pilot-length/active-device-number plane or some other parameter space would improve clarity.
- [Abstract] The term 'massive grant-free access' could be made more precise by giving example numbers for the number of devices, the number of active devices, and the pilot length.
Circularity Check
No circularity found in the abstract; the derivation is self-contained and no fitted input is renamed as prediction.
full rationale
The abstract-only manuscript presents a time-domain signal model, formulates MAP/MMSE estimation problems, builds a factor graph from the model statistics, and proposes AMP-based algorithms to approximately solve those problems. No step in the abstract reduces to its own input by definition: the factor graph is derived from the stated channel and activity statistics, and the algorithms are proposed as approximate solvers, not as fitted outputs. There is also no visible self-citation chain or imported uniqueness theorem, and no fitted parameter is later called a prediction. State evolution analysis and computational complexity claims are presented as analytic results under the stated model, not as consequences of assuming the conclusion. Since no specific equation-level reduction or predict-then-fit pattern can be quoted from the available text, the appropriate finding is no significant circularity.
Assumptions & free parameters
assumptions (1)
- domain assumption The time-domain channel and device activity statistics are exactly known and captured by the new factor graph.
Cite this review
Pith. "Pith review of AMP-based Joint Activity Detection and Channel Estimation for Massive Grant-Free Access in OFDM-based Wideband Systems." pith.science (2026). https://pith.science/paper/K6JGDIIU
@misc{pith2026250806540,
author = {Pith},
title = {Pith review of: AMP-based Joint Activity Detection and Channel Estimation for Massive Grant-Free Access in OFDM-based Wideband Systems},
year = {2026},
howpublished = {\url{https://pith.science/paper/K6JGDIIU}},
note = {Machine review of arXiv:2508.06540}
}
read the original abstract
To realize orthogonal frequency division multiplexing (OFDM)-based grant-free access for wideband systems under frequency-selective fading, existing device activity detection and channel estimation methods need substantial accuracy improvement or computation time reduction. In this paper, we aim to resolve this issue. First, we present an exact time-domain signal model for OFDM-based grant-free access under frequency-selective fading. Then, we present a maximum a posteriori (MAP)-based device activity detection problem and two minimum mean square error (MMSE)-based channel estimation problems. The MAP-based device activity detection problem and one of the MMSE-based channel estimation problems are formulated for the first time. Next, we build a new factor graph that captures the exact statistics of time-domain channels and device activities. Based on it, we propose two approximate message passing (AMP)-based algorithms, AMP-A-EC and AMP-A-AC, to approximately solve the MAP-based device activity detection problem and two MMSE-based channel estimation problems. Both proposed algorithms alleviate the AMP's inherent convergence problem when the pilot length is smaller or comparable to the number of active devices. Then, we analyze AMP-A-EC's error probability of activity detection and mean square error (MSE) of channel estimation via state evolution and show that AMP-A-AC has the lower computational complexity (in dominant term). Finally, numerical results show the two proposed AMP-based algorithms' superior performance and respective preferable regions, revealing their significant values for OFDM-based grant-free access.
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.