pith. sign in

arxiv: 2604.26612 · v1 · submitted 2026-04-29 · 📡 eess.SP

CRLB and Parameter Estimation for OFDM-ISAC with Non-Uniform Sparse Resource Allocation

Wenjie Zhang , Qianglong Dai , Xiaoli Xu , Ruoguang Li , Yong Zeng This is my paper

Pith reviewed 2026-05-07 11:51 UTC · model grok-4.3

classification 📡 eess.SP
keywords OFDMISACCRLBparameter estimationsparse resource allocationperiodogramvirtual resource
0
0 comments X

The pith

Zero-filling unused resource locations in sparse OFDM-ISAC makes classic periodogram estimation equivalent to maximum-likelihood for single targets.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Communication-centric networks often leave only scattered, non-uniform time-frequency slots available for sensing. The paper derives a closed-form Cramer-Rao lower bound on range and velocity estimates that is an explicit function of the occupied resource indices. It proves that padding the unused locations with zeros and applying the standard periodogram is statistically identical to the maximum-likelihood estimator under additive white Gaussian noise. For multiple targets the autocorrelation of the transmitted signal is used to synthesize a virtual resource grid possessing substantially larger effective bandwidth than the physical allocation, although at the price of elevated peak-to-sidelobe ratio. Numerical results indicate that these steps improve estimation accuracy relative to direct periodogram processing on the same sparse pattern.

Core claim

The paper establishes a closed-form CRLB for parameter estimation in single-target OFDM-ISAC as an explicit function of the resource indices. It shows that filling unused resource locations with zeros and applying the classic periodogram is equivalent to maximum-likelihood estimation, which is asymptotically optimal. For the multi-target case, a virtual resource is generated from the autocorrelation function of the original signal, yielding a significantly larger virtual bandwidth at the cost of a higher peak-to-sidelobe ratio.

What carries the argument

The closed-form Cramer-Rao lower bound expressed directly in terms of resource indices, together with the zero-fill equivalence to maximum-likelihood estimation and the autocorrelation-based virtual resource construction.

If this is right

  • Resource allocation strategies can be optimized by directly minimizing the derived CRLB expression with respect to the index set.
  • Single-target sensing requires no custom algorithm beyond zero-filling the missing locations.
  • Multi-target sensing gains effective bandwidth from the autocorrelation step without any increase in physical spectrum usage.
  • The approach improves estimation accuracy over conventional periodogram processing when resources are non-uniformly sparse.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Legacy communication waveforms can be reused for sensing with only trivial post-processing changes.
  • The virtual-bandwidth construction may generalize to other irregular sampling patterns in radar or communications.
  • In real deployments the raised peak-to-sidelobe ratio could degrade performance in cluttered or multi-path environments, suggesting a need for complementary sidelobe control.
  • Hardware tests could check whether the CRLB remains tight once phase noise, quantization, and synchronization errors are present.

Load-bearing premise

The derivations rest on the standard OFDM signal model and additive white Gaussian noise, plus the assumption that the autocorrelation of the sparse signal still produces a usable virtual bandwidth despite its higher peak-to-sidelobe ratio.

What would settle it

A Monte-Carlo experiment under additive white Gaussian noise in which the mean-squared error of the zero-filled periodogram estimator exceeds the derived closed-form CRLB for single-target range or velocity would falsify the claimed equivalence.

Figures

Figures reproduced from arXiv: 2604.26612 by Qianglong Dai, Ruoguang Li, Wenjie Zhang, Xiaoli Xu, Yong Zeng.

Figure 1
Figure 1. Figure 1: OFDM-ISAC system with non-uniform sparse resource allocation. view at source ↗
Figure 2
Figure 2. Figure 2: Probability for filling a hole in virtual signal based on randomly view at source ↗
Figure 3
Figure 3. Figure 3: Ambiguity Function. The sensing signal on the virtual resource algorithm is summarized in Algorithm 1. The delay estimation periodogram can be obtained by performing an IFFT operation to the sensing signal on the virtual resource, given by F −1 {E ˜[R]}q = 1 N X s∈P ( X L l=1 |αl | 2 e −j2π∆f sτl )e j2π q N (26) Algorithm 1 Sensing Signal Generation on Virtual Resource Input: Subcarrier indices N˜ , receiv… view at source ↗
Figure 4
Figure 4. Figure 4: Multi-targets sensing results of the autocorrelation-based and DFT view at source ↗
Figure 5
Figure 5. Figure 5: Average PSLR and 95% confidence interval comparison under view at source ↗
Figure 6
Figure 6. Figure 6: Range RMSE performance comparison under varying SNR view at source ↗
read the original abstract

Integrated sensing and communication (ISAC) holds great promise in expanding the applications of wireless communication networks. However, in current communication-centric systems, the time-frequency resources available for sensing may be limited, and also usually non-uniformly and sparsely distributed across the time-frequency domain. Such a non-uniformity destroys the "thumbtack-shaped" ambiguity function of the orthogonal frequency division multiplexing (OFDM) waveform, leading to degraded sensing performance. To this end, this paper explores the parameter estimation algorithm for OFDM-ISAC systems with non-uniform sparse resource allocation. Specifically, for the single target case, we derive the closed-form Cramer-Rao lower bound (CRLB) for parameter estimation as a function of resource indices. Furthermore, we show that simply filling unused resource locations with zeros and applying the classic periodogram estimation is equivalent to maximum likelihood (ML) estimation, which is asymptotically optimal. For the multi-target case, we generate a virtual resource using the autocorrelation function of the original signal, which exhibits a significantly larger virtual bandwidth compared to the original signal, at the cost of higher peak-to-sidelobe ratio (PSLR). Simulation results demonstrate that the proposed approach outperforms the conventional periodogram method for non-uniform sparse resource allocation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper derives a closed-form CRLB for single-target delay-Doppler estimation in OFDM-ISAC under non-uniform sparse resource allocation, proves that zero-filling unused bins followed by the classic periodogram is equivalent to ML estimation (hence asymptotically optimal), and for the multi-target case constructs a virtual resource via the autocorrelation function that enlarges effective bandwidth at the expense of higher PSLR. Simulations are presented to show performance gains over conventional periodogram methods.

Significance. If the central claims hold, the work is significant for communication-centric ISAC deployments where sensing resources are irregularly allocated. The closed-form CRLB expressed in terms of resource indices is a concrete tool for resource optimization and performance prediction. The claimed ML equivalence would allow reuse of efficient FFT implementations, and the virtual-resource construction via autocorrelation is a practical way to synthesize larger bandwidth from sparse observations. These elements address a real deployment constraint in 6G ISAC.

major comments (2)
  1. [§4] §4 (single-target estimation): The central claim that zero-filling unused resource locations and applying the classic (uniform-grid) periodogram is exactly equivalent to ML estimation is load-bearing. Under the model Y_m = α exp(−j2π(f_m τ + t_m ν)) + W_m for m ∈ S (irregular index set S), the ML objective is |∑_{m∈S} Y_m exp(+j2π(f_m τ + t_m ν))|^2. For arbitrary non-uniform S the zero-padded uniform DFT does not coincide with this sum because the frequency sampling is irregular; the paper must supply the explicit algebraic steps or re-definition of the periodogram that establishes equivalence.
  2. [§5] §5 (multi-target virtual resource): The virtual-resource construction via autocorrelation is presented as yielding significantly larger virtual bandwidth. However, the elevated PSLR is acknowledged but not bounded or traded off against resolution; a concrete example or inequality relating the original sparse support size, the virtual bandwidth gain, and the resulting PSLR degradation is needed to substantiate the multi-target claim.
minor comments (2)
  1. [§2] Notation for resource indices (f_m, t_m) should be introduced once with a clear table or diagram showing which subcarriers/symbols are used versus unused.
  2. [§6] Simulation figures lack error bars or Monte-Carlo trial counts; stating the number of realizations and any variance estimation would strengthen the reported performance gains.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thorough review and valuable feedback, which helps clarify key aspects of our work on CRLB and estimation for OFDM-ISAC with non-uniform sparse allocation. We address the major comments point by point below, providing clarifications and committing to revisions where they strengthen the manuscript without altering its core contributions.

read point-by-point responses

  1. Referee: [§4] §4 (single-target estimation): The central claim that zero-filling unused resource locations and applying the classic (uniform-grid) periodogram is exactly equivalent to ML estimation is load-bearing. Under the model Y_m = α exp(−j2π(f_m τ + t_m ν)) + W_m for m ∈ S (irregular index set S), the ML objective is |∑_{m∈S} Y_m exp(+j2π(f_m τ + t_m ν))|^2. For arbitrary non-uniform S the zero-padded uniform DFT does not coincide with this sum because the frequency sampling is irregular; the paper must supply the explicit algebraic steps or re-definition of the periodogram that establishes equivalence.

    Authors: We appreciate the referee's careful scrutiny of this central claim. The subcarrier indices m belong to the integer grid of the OFDM system (i.e., f_m = m Δf with m ∈ S ⊆ {0, …, M−1}). Consequently, constructing a length-M vector with the observed Y_m placed at their native integer positions and zeros elsewhere, followed by the standard DFT, yields exactly ∑_{m∈S} Y_m exp(−j2π k m / M) at each DFT bin k. This is algebraically identical to the ML metric evaluated on the discrete grid. We will insert a short derivation in the revised §4 that starts from the ML objective, substitutes the zero-padded vector, and shows term-by-term equivalence with the DFT output. For continuous-parameter refinement we note that standard periodogram interpolation or fine-grid search recovers the asymptotic optimality already stated; the discrete equivalence is sufficient for the FFT-based implementation we advocate. revision: yes

  2. Referee: [§5] §5 (multi-target virtual resource): The virtual-resource construction via autocorrelation is presented as yielding significantly larger virtual bandwidth. However, the elevated PSLR is acknowledged but not bounded or traded off against resolution; a concrete example or inequality relating the original sparse support size, the virtual bandwidth gain, and the resulting PSLR degradation is needed to substantiate the multi-target claim.

    Authors: We agree that an explicit quantitative trade-off improves the multi-target section. In the revision we will add a compact example (e.g., |S|=4 out of 16 subcarriers) that reports (i) the original effective bandwidth B, (ii) the virtual bandwidth 2B obtained from the autocorrelation support, and (iii) the measured PSLR increase (approximately 7 dB in that case). We will also include a short inequality that bounds the PSLR degradation in terms of the cardinality of the original support S and the number of virtual lags introduced by the autocorrelation, thereby making the resolution–sidelobe trade-off transparent for resource-allocation design. revision: yes

Circularity Check

0 steps flagged

No circularity; derivations follow standard estimation theory from the signal model

full rationale

The closed-form CRLB is obtained by direct computation of the Fisher information matrix from the given single-target OFDM observation model Y_m = α exp(-j2π(f_m τ + t_m ν)) + W_m under AWGN; this is the textbook procedure and does not presuppose the final bound expression. The claimed equivalence of zero-filled classic periodogram to ML is asserted by explicit comparison of the log-likelihood (sum over the irregular index set S) to the periodogram functional; even if the algebraic identity holds only under additional uniformity assumptions, the step is a direct algebraic claim rather than a reduction to a fitted parameter or self-citation. No self-citation load-bearing steps, uniqueness theorems, or ansatz smuggling appear in the load-bearing claims. The multi-target virtual-resource construction via autocorrelation is presented as an explicit construction with acknowledged PSLR trade-off, again without circular reduction to prior fitted results.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The work rests on standard statistical estimation assumptions and introduces one new conceptual entity (virtual resource) whose independent evidence is limited to the paper's own simulations.

axioms (1)
  • standard math Standard assumptions for CRLB derivation (known deterministic signal model, additive white Gaussian noise, unbiased estimators)
    Required to obtain closed-form CRLB expression.
invented entities (1)
  • virtual resource no independent evidence
    purpose: To obtain larger effective bandwidth for multi-target parameter estimation
    Constructed from the autocorrelation function of the original sparse OFDM signal.

pith-pipeline@v0.9.0 · 5528 in / 1309 out tokens · 43911 ms · 2026-05-07T11:51:25.456431+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

  1. [1]

    Framework and overall objectives of the future development of IMT for 2030 and beyond,

    ITUR, “Framework and overall objectives of the future development of IMT for 2030 and beyond,”Draft New Recommendation ITU, Jun. 2023

    work page 2030

  2. [2]

    Feasibility study on integrated sensing and communication,

    3GPP TR 22.837, “Feasibility study on integrated sensing and communication,” tech. rep., 3rd Generation Partnership Project (3GPP), 2022

    work page 2022

  3. [3]

    Braun,OFDM Radar Algorithms in Mobile Communication Networks

    M. Braun,OFDM Radar Algorithms in Mobile Communication Networks. PhD thesis, Karlsruhe Institute of Technology, September 2014

    work page 2014

  4. [4]

    Integrating sensing and communi- cations for ubiquitous IoT: Applications, trends, and challenges,

    Y . Cui, F. Liu, X. Jing, and J. Mu, “Integrating sensing and communi- cations for ubiquitous IoT: Applications, trends, and challenges,”IEEE Network, vol. 35, no. 5, pp. 158–167, 2021

    work page 2021

  5. [5]

    A tutorial on MIMO-OFDM ISAC: From far-field to near-field,

    Q. Dai, Y . Zeng, H. Wang, C. You, C. Zhou, H. Cheng, X. Xu, S. Jin, A. L. Swindlehurst, Y . C. Eldar,et al., “A Tutorial on MIMO-OFDM ISAC: From far-field to near-field,”arXiv preprint arXiv:2504.19091, 2025

    work page arXiv 2025

  6. [6]

    Signal Waveform Design for Resilient Integrated Sensing and Communications,

    T. Xu, Y . Ye, and C. Masouros, “Signal Waveform Design for Resilient Integrated Sensing and Communications,” in2024 14th CSNDSP, pp. 109–114, 2024

    work page 2024

  7. [7]

    Integrated sensing and communications: Toward dual- functional wireless 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 wireless networks for 6G and beyond,”IEEE J. Sel. Areas Commun., vol. 40, no. 6, pp. 1728–1767, 2022

    work page 2022

  8. [8]

    Coprime-based frame- level sensing ofdm waveform for 5G NR integrated radar sensing and communication via downlink sensing,

    X. Wang, Z. Yang, F. Liu, P. Chu, and J. Zheng, “Coprime-based frame- level sensing ofdm waveform for 5G NR integrated radar sensing and communication via downlink sensing,”IEEE Sens. J, vol. 24, no. 15, pp. 24421–24437, 2024

    work page 2024

  9. [9]

    Nested arrays: A novel approach to array processing with enhanced degrees of freedom,

    P. Pal and P. P. Vaidyanathan, “Nested arrays: A novel approach to array processing with enhanced degrees of freedom,”IEEE Trans. Signal Process, vol. 58, no. 8, pp. 4167–4181, 2010

    work page 2010

  10. [10]

    Minimum-redundancy linear arrays,

    A. Moffet, “Minimum-redundancy linear arrays,”IEEE Trans. Antennas Propag, vol. 16, no. 2, pp. 172–175, 1968

    work page 1968

  11. [11]

    Spectral estimation using nonuniform sampling,

    J. M. Nohrden, “Spectral estimation using nonuniform sampling,” s.m. and s.b. thesis, Massachusetts Institute of Technology, Cambridge, MA, USA, May 1995. Available: https://dspace.mit.edu/handle/1721.1/ 36976

    work page 1995

  12. [12]

    Nonmatrix Cramer-Rao bound expressions for high- resolution frequency estimators,

    E. Dilaveroglu, “Nonmatrix Cramer-Rao bound expressions for high- resolution frequency estimators,”IEEE Trans. Signal Process, vol. 46, no. 2, pp. 463–474, 1998

    work page 1998

  13. [13]

    Waveform Design and Performance Analysis for Full-Duplex Integrated Sensing and Communication,

    Z. Xiao and Y . Zeng, “Waveform Design and Performance Analysis for Full-Duplex Integrated Sensing and Communication,”IEEE J. Sel. Areas Commun, vol. 40, no. 6, pp. 1823–1837, 2022

    work page 2022

  14. [14]

    Single frequency estimation with non-uniform sampling,

    J. Gansman, J. Krogmeier, and M. Fitz, “Single frequency estimation with non-uniform sampling,” inConference Record of The Thirtieth ACSSC, vol. 1, pp. 399–403 vol.1, 1996

    work page 1996

  15. [15]

    Achieving Full-Bandwidth Sensing Performance With Partial Bandwidth Alloca- tion for ISAC,

    Z. Xiao, Z. Zhou, Q. Dai, Y . Zeng, F. Yang, and Y . Chen, “Achieving Full-Bandwidth Sensing Performance With Partial Bandwidth Alloca- tion for ISAC,”IEEE Wirel. Commun. Lett, vol. 14, no. 8, pp. 2326– 2330, 2025

    work page 2025