Pith. sign in

REVIEW 5 major objections 5 minor 1 cited by

Target Detection in OFDM-ISAC Systems: A Multipath Exploitation Approach

T0 review · 5 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper argues that multipath propagation, normally a nuisance in radar sensing, can be exploited as diversity: a weighted generalized likelihood ratio test on OFDM echo data, with jointly optimized subcarrier powers and detector…

desk verdict The multipath diversity idea is worth a look, but the paper's weight solution and false-alarm invariance rest on false equalities and an unverified orthogonality assumption, so the claimed gains are not established. read the letter →

arxiv 2501.07893 v1 pith:O5GKSCLC submitted 2025-01-14 eess.SP

Xiaohan Lv , Rang Liu , Ming Li , Qian liu This is my paper · ORCID
classification eess.SP
keywords integratedsensingandcommunicationOFDMmultipathexploitationtargetdetectiongeneralizedlikelihoodratiotestdelay-Dopplerdiversitypowerallocationmajorization-minimization
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

Most radar sensing treats multipath echoes as clutter or ambiguity; this paper argues that in an OFDM system that communicates and senses from the same waveform, the reflected paths between base station and target can be harvested as extra diversity. The authors develop a weighted generalized likelihood ratio test (a statistical decision rule for target present versus absent) that combines echoes from all resolvable paths, each with its own delay and Doppler shift. They then jointly design the subcarrier transmit powers and the per-path detector weights, maximizing detection probability subject to the communication user's SNR requirement and a total power budget. Simulation results show that the joint design outperforms a conventional detector that uses only the line-of-sight echo, and that the gain grows when the non-line-of-sight reflections carry more energy.

What carries the argument

The central object is the weighted GLRT statistic built from rank-one projection matrices $P_{s,l} = \tilde{\mathbf{H}}_l(\tilde{\mathbf{H}}_l^H \tilde{\mathbf{H}}_l)^{\dagger} \tilde{\mathbf{H}}_l^H$ onto each path's delay-Doppler channel vector. The load-bearing property is that these projections are mutually orthogonal, which lets the authors apply the Craig-Sakamoto theorem and write the detector as a ratio of independent F-distributed pieces; this is what makes the false-alarm distribution depend on only $N$, $M$, and $L$. The optimization machinery is an alternating loop: with weights fixed, the non-concave power-allocation objective is replaced by a concave first-order surrogate (the majorization-minimization step), and with power fixed, the weight vector is the normalized vector of path strengths, $w = d/\|d\|_2$, which solves a Rayleigh quotient.

What would settle it

Run the detector in a simulated two-path scenario in which the two paths land on the same discretized delay tap (or the same Doppler tap), so the projection matrices are no longer orthogonal, and measure the empirical false-alarm probability for several different power-allocation and weight choices; if the measured $P_{\mathrm{FA}}$ shifts with the design variables, the central claim that the threshold depends only on $N$, $M$, and $L$ is falsified.

Watch

Extended reading notes

Core claim

The central claim is that a GLRT detector on multipath OFDM echoes decomposes into a sum of per-path projection terms, so that detection can be improved by weighting those terms and optimizing transmit power. Specifically, with $L$ separable delay-Doppler paths, the weighted statistic is $\tilde{\eta} = 1 + \sum_{l=1}^{L} w_l^2 \|P_{s,l}Y\|_F^2 / \|P_n Y\|_F^2$, where $P_{s,l}$ projects onto the $l$-th path's channel vector and $P_n$ onto the complementary null space. The paper derives that under the no-target hypothesis this statistic follows an F-distribution determined only by $N$, $M$, and $L$, independent of the transmit power allocation and the weights, so the false-alarm threshold does not move during design. It then maximizes the non-centrality parameter $\sum_l w_l^2 \|H_l A X\|_F^2$ in place of the unknown RCS-weighted term, and shows by simulation that the resulting joint power-and-weight design achieves materially better ROC performance than the LoS-only conventional detector.

Load-bearing premise

The argument depends on the discrete delay-Doppler channel vectors of different paths being exactly orthogonal, so the per-path projection matrices decouple and the false-alarm distribution is F; this exact orthogonality holds only on an idealized grid of distinct integer delay and Doppler taps, and the paper does not test how performance changes when real channels leave that grid.

Editorial extensions

If this is right

  • In a multipath-rich cell, an OFDM-ISAC base station can detect targets whose direct line-of-sight echo is weak by pooling non-line-of-sight reflections.
  • The false-alarm threshold can be set once from the system dimensions $N$, $M$, and $L$, so transmit power and detector weights can be optimized without re-tuning the detection threshold.
  • Jointly optimizing subcarrier powers and detector weights yields better ROC performance than optimizing either one alone, and the improvement is driven mainly by the weight design.
  • The optimized detector removes the ghost targets that the unoptimized detector produces on the range-Doppler grid.
  • Communication service is protected during the optimization because every subcarrier's SNR at the user is kept above the required level.

Reading between the lines

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

  • Because the false-alarm invariance relies on exact projection orthogonality, a natural robustness test would be to simulate fractional or clustered delay/Doppler taps and watch whether $P_{\mathrm{FA}}$ stays flat; the paper does not report such a test.
  • The same weighted-combining idea likely extends to MIMO-OFDM ISAC, where multiple transmit and receive antennas provide additional diversity branches beyond the delay-Doppler paths, though the paper does not analyze that case.
  • The optimization objective drops the RCS coefficients $\Lambda_l$ and uses path-strength surrogates, so the weights favor geometrically strong paths; a stochastic version that accounts for path-dependent RCS fluctuations could behave differently under Swerling-type targets.
  • If the communication symbols $x_{n,m}$ are random rather than known pilots, the power-allocation objective would need to be averaged over the data; the paper treats $X$ as given, leaving a data-dependent extension open.
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

5 major / 5 minor

Summary. This paper proposes a weighted GLRT detector for target detection in an OFDM-based ISAC system under multipath propagation, together with a joint design of subcarrier power allocation and detector weights via an MM-based alternating algorithm. The main theoretical claims are: (i) the detection statistic under H0 follows a parameter-free F-distribution independent of the power allocation and weights; (ii) the objective f(A,w) = sum_l w_l^2 ||H_l A X||^2_F is positively proportional to the detection probability; and (iii) the proposed design achieves significantly better target detection performance than the conventional detector, as supported by ROC simulations in Section V.

Significance. The problem of exploiting multipath for sensing in ISAC systems is relevant and timely, and the paper formulates a clean system model. If the theoretical results were correct, the proposed low-complexity joint design would be a useful contribution. However, the derivations contain multiple load-bearing errors: the GLRT statistic is not derived correctly, the false-alarm distribution claim is false, the weight design solves a different optimization problem than stated, and the proposed objective is not tied to detection probability. The simulation results cannot repair these gaps because the thresholds and comparisons rely on the invalid analysis. The claimed significance is therefore not established.

major comments (5)
  1. [Section III, Eq. (12)] The reformulation of the GLRT as eta = ||Y||^2_F / ||Y - sum_l P_{s,l}Y||^2_F assumes that hat(Lambda)_l H_l A X = P_{s,l} Y for each l. From the given definition hat(Lambda)_l = P_{s,l} Y X^dagger A^dagger H_l^dagger, one obtains hat(Lambda)_l H_l A X = P_{s,l} Y (X^dagger A^dagger H_l^dagger H_l A X), and the factor in parentheses is not generally an identity. Since X is a block-diagonal symbol matrix that is not square and not isometric, and H_l A X differs from tilde(H)_l, the equality is not generally valid. This invalidates the derivation of the GLRT statistic on which the entire detector design is based.
  2. [Section III, Eq. (15)] The claim that the false-alarm statistic tilde(eta)_0 follows an F-distribution whose parameters depend only on N, M, and L is not correct. Under H0 with white Gaussian noise, the numerator equals ||sum_l w_l P_{s,l}Z||^2_F = sum_l w_l^2 ||P_{s,l}Z||^2_F, which is a weighted sum of independent chi-square variables with weights w_l^2; it is a chi-square only if all nonzero w_l^2 are equal. The ratio therefore does not have the stated central F-distribution, and the false-alarm probability depends on w. The constant false-alarm rate property asserted in Section III is thus unsupported, and the simulation thresholds in Section V are not justified.
  3. [Section III, after Eq. (11)] The derivation assumes exact mutual orthogonality of the channel matrices of different paths, i.e., tilde(H)_{l1}^H tilde(H)_{l2} = 0 for l1 != l2. This holds only if the discretized delay and Doppler tap indices are distinct integers modulo N and M, respectively, which is neither stated nor verified. The continuous delay/Doppler model in (4) and the rounding in (5) do not guarantee this condition for the simulation parameters in Table I. Without exact orthogonality, the Craig-Sakamoto simplification in (13), the equivalence in (15), and the non-centrality parameter simplification in (16) all fail.
  4. [Section IV-B, Eq. (23a)] The equality sum_l w_l^2 ||H_l A X||^2_F = |w^H d|^2 is algebraically false: the left-hand side is sum_l w_l^2 d_l^2, whereas the right-hand side equals sum_l w_l^2 d_l^2 + 2 sum_{l<k} w_l w_k d_l d_k. For nonnegative w and ||w||^2=1, maximization of sum_l w_l^2 d_l^2 is solved by a unit vector at the largest d_l, not by w = d/||d||. Consequently the closed-form solution in (24) and the alternating algorithm built on it do not solve the stated problem (18).
  5. [Section III, Eq. (17)] The statement that f(A,w) is 'positively proportional to the detection probability' is asserted without proof. The actual non-centrality parameter under H1 involves the random RCS matrices Lambda_l, which are omitted from f(A,w). Moreover, because the false-alarm threshold depends on w, maximizing f(A,w) does not necessarily maximize the detection probability at a fixed false-alarm rate. Thus problem (18) is not aligned with its stated objective.
minor comments (5)
  1. [Section II, after Eq. (4)] The definitions of the tap indices are swapped: the expression k_l = round[nu_l M T] cannot be a delay tap index and r_l = round[tau_l N Delta f] cannot be a Doppler tap index; they should be k_l = round[tau_l N Delta f] and r_l = round[nu_l M T].
  2. [Eq. (1)] The exponential e^{j2*pi*(f0 + n*Delta*f*(t - m*T))} is ambiguous; it should be e^{j2*pi*(f0 + n*Delta*f)*(t - m*T)}.
  3. [Section IV-A, after Eq. (20)] The sentence 'Now the objective function is now expressed as a quadratic form' contains a duplicated 'now'; the text should be revised for clarity.
  4. [Section V] The text says 'Swelling-I type' RCS; the standard radar term is 'Swerling-I'. Also, Table I presents the beta values in an unclear format; it should specify which value corresponds to which path.
  5. [Section V] The paper does not state how the threshold xi in (14) is chosen in the simulations; this is important because the analysis claims a parameter-free false-alarm distribution, and the simulations should verify this claim directly.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reduction found: the weighted GLRT derivation and the joint power/weight design are self-contained, with unproved modeling assumptions (exact path orthogonality and the use of f as a detection-probability proxy) that are correctness risks rather than circularity.

full rationale

The paper does not fit parameters to data and then predict the same data; no fitted constant is renamed a prediction. The derivation chain is: (11) is the standard GLRT; (12)-(13) use the projection identity P_{s,l}Y and Craig-Sakamoto under the stated orthogonality P_{s,l1}P_{s,l2}^H = 0. That orthogonality is an assumption on the discrete delay-Doppler grid, not a result imported from the authors' prior work, and it does not make the output equivalent to the input. The weighted detector (14) and the false-alarm distribution (15) rest on the same orthogonality; whether the condition actually holds for the simulated continuous/off-grid taps is a mathematical-validity question, not a circularity. The optimization objective f(A,w) in (17) is asserted to be 'positively proportional to the detection probability'; this is an unproved surrogate, but the paper does not define f in terms of the measured detection probability or fit f to detection outcomes. It is an ansatz, not a self-referential reduction. Self-citations are limited to a general ISAC reference in the introduction and are not load-bearing; no uniqueness theorem from the same authors is invoked. Accordingly, no step qualifies as circular under the quoted-reduction standard. The honest finding is no circularity, with the caveat that the surrogate-objective and exact-orthogonality claims carry correctness risk that should be addressed by a numerical orthogonality check and independent Monte Carlo replication.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no fitted constants, but it rests on several unstated modeling assumptions: exact path orthogonality on the discrete grid, a channel-only projection subspace for the GLRT, and a claim that dropping the RCS coefficients preserves proportionality to detection probability. The weight update also uses a false algebraic identity.

assumptions (3)
  • domain assumption Discrete delay-Doppler channel vectors are orthogonal across paths: ~H_{l1} ~H_{l2}^H = 0 for l1 != l2.
    Used in Section III to apply the Craig-Sakamoto theorem and simplify the GLRT (13). Only exact on an ideal grid with equal numbers of subcarriers and symbols and distinct integer delay/Doppler taps.
  • ad hoc to paper The signal subspace for each path is spanned by the channel vectors ~H_l, independent of the transmitted symbols and power allocation.
    The projection P_{s,l} is defined from ~H_l only, so the detector ignores the modulation structure in H_l A X. This is not the GLRT for the stated model.
  • ad hoc to paper The non-centrality parameter is proportional to sum_l w_l^2 ||H_l A X||^2_F, with the RCS coefficients Lambda_l omitted.
    The paper asserts f(A,w) is positively proportional to detection probability after (16), but the true non-centrality contains ||Lambda_l H_l A X||^2, so the proportionality is not established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Target Detection in OFDM-ISAC Systems: A Multipath Exploitation Approach." pith.science (2026). https://pith.science/paper/O5GKSCLC

@misc{pith2026250107893,
  author       = {Pith},
  title        = {Pith review of: Target Detection in OFDM-ISAC Systems: A Multipath Exploitation Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/O5GKSCLC}},
  note         = {Machine review of arXiv:2501.07893}
}
read the original abstract

This paper investigates the potential of multipath exploitation for enhancing target detection in orthogonal frequency division multiplexing (OFDM)-based integrated sensing and communication (ISAC) systems. The study aims to improve target detection performance by harnessing the diversity gain in the delay-Doppler domain. We propose a weighted generalized likelihood ratio test (GLRT) detector that effectively leverages the multipath propagation between the base station (BS) and the target. To further enhance detection accuracy, a joint optimization framework is developed for subcarrier power allocation at the transmitter and weight coefficients of the GLRT detector. The objective is to maximize the probability of target detection while satisfying constraints on total transmit power and the communication receiver's signal-to-noise ratio (SNR). An iterative algorithm based on the majorization-minimization (MM) method is employed to address the resulting non-convex optimization problem. Simulation results demonstrate the efficacy of the proposed algorithm and confirm the benefits of multipath exploitation for target detection in OFDM-ISAC systems under multipath-rich environments.

Figures

Figures reproduced from arXiv: 2501.07893 by the authors.

Figure 1
Figure 1. An ISAC system under multipath propagation scenario [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 4
Figure 4. Two-dimensional delay-Doppler results (the target [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 3
Figure 3. Probability of detection versus the RCS variance of [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Beyond Single-Band: Analysis and Resource Allocation for Multi-band ISAC Systems

    eess.SP 2026-07 conditional novelty 6.0 of 10

    Closed-form detection probabilities for multi-band OFDM ISAC signals are derived via characteristic functions of i.n.i.d. exponential variables, and an ADMM-based resource allocator achieves 18 dB detection gain over ...

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages · cited by 1 Pith paper

  1. [5]

    Zhang, M

    A. Zhang, M. L. Rahman, X. Huang, Y. J. Guo, S. Chen, and R. W. Heath, ``Perceptive mobile networks: Cellular networks with radio vision via joint communication and radar sensing,'' IEEE Veh. Technol. Mag., vol. 16, no. 2, pp. 20-30, Jun. 2021

    work page 2021

  2. [1]

    M. Z. Chowdhury, M. Shahjalal, S. Ahmed and Y. M. Jang, ``6G wireless communication systems: Applications, requirements, technologies, challenges, and research directions,'' IEEE Open J. Commun. Soc. , vol. 1, pp. 957-975, 2020

    work page 2020

  3. [2]

    F. Liu, Y. Cui, C. Masouros, J. Xu, T. X. Han, Y. C. Eldar, and S. Buzz, ``Integrated sensing and communications: Toward dual-functional wireless networks for 6G and beyond,'' IEEE J. Sel. Areas Commun., vol. 40, no. 6, pp. 1728-1767, Jun. 2022

    work page 2022

  4. [3]

    R. Liu, M. Li, Y. Liu, Q. Wu, and Q. Liu, ``Joint transmit waveform and passive beamforming design for RIS-aided DFRC systems," IEEE J. Sel. Topics Signal Process., vol. 16, no. 5, pp. 995-1010, Aug. 2022

    work page 2022

  5. [4]

    Mercier, S

    S. Mercier, S. Bidon, D. Roque, and C. Enderli, ``Comparison of correlation-based OFDM radar receivers,'' IEEE Trans. Aerosp. Electron. Syst. , vol. 56, no. 6, pp. 4796-4813, Dec. 2020

    work page 2020

  6. [6]

    Leigsnering, F

    M. Leigsnering, F. Ahmad, M. G. Amin, and A. M. Zoubir, ``Multipath exploitation in sparse scene recovery using sensing-through-wall distributed radar sensor configurations,'' in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP), South Brisbane, Australia, Apr. 2015, pp. 2749-2753

    work page 2015

  7. [7]

    Sen and A

    S. Sen and A. Nehorai, ``Adaptive OFDM radar for target detection in multipath scenarios'', IEEE Trans. Signal Process., vol. 59, no. 1, pp. 78-90, Jan. 2011

    work page 2011

  8. [8]

    Z. Xu, C. Fan, and X. Huang, ``MIMO radar waveform design for multipath exploitation,'' IEEE Trans. Signal Process., vol. 69, pp. 5359-5371, Oct. 2021

    work page 2021

Show all 15 references
  1. [9]

    Rihan, E

    M. Rihan, E. Grossi, L. Venturino, and S. Buzzi, ``Spatial diversity in radar detection via active reconfigurable intelligent surfaces,'' IEEE Signal Process. Lett., vol. 29, pp. 1242-1246, May 2022

    2022

  2. [10]

    Wiesbeck and L

    W. Wiesbeck and L. Sit, ``Radar 2020: The future of radar systems,'' in Proc. Int. Radar Conf. , Lille, France, Oct. 2014, pp. 1-6

    2020

  3. [11]

    Hadani, S

    R. Hadani, S. Rakib, M. Tsatsanis, A. Monk, A. J. Goldsmith, A. F. Molisch, and R. Calderbank, ``Orthogonal time frequency space modulation,'' in Proc. IEEE Wireless Commun. Netw. Conf. (WCNC), San Francisco, CA, Apr. 2017, pp. 1-6

    2017

  4. [12]

    S. M. Kay, Fundamentals of Statistical Signal Processing: Detection Theory, vol. 2. Englewood Cliffs, NJ, USA: Prentice-Hall, 1998

    1998

  5. [13]

    Dogandzic and A

    A. Dogandzic and A. Nehorai, ``Generalized multivariate analysis of variance: A unified framework for signal processing in correlated noise,'' IEEE Signal Process. Mag., vol. 20, pp. 39-54, Sep. 2003

    2003

  6. [14]

    Letac and H

    G. Letac and H. Massam, ``Craig-Sakamoto's theorem for the Wishart distributions on symmetric cones,'' Ann. Inst. Stat. Math., vol. 47, pp. 785-799, Feb. 1995

    1995

  7. [15]

    Y. Sun, P. Babu, and D. P. Palomar, ``Majorization-minimization algorithms in signal processing communications and machine learning,'' IEEE Trans. Signal Process. , vol. 65, no. 3, pp. 794-816, Feb. 2017

    2017

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.