Pith. sign in

REVIEW 3 major objections 4 minor 10 references

A Smooshed BMOCZ Zero Constellation for CFO Estimation Without Channel Coding

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Smooshed zero constellation lets a receiver estimate and correct carrier-frequency-offset rotation by finding a gap in the received polynomial, with no channel coding.

desk verdict A genuinely new CFO-estimation idea for BMOCZ, but the paper's real weak spot is an unproven peak-uniqueness claim; the reviewer's main objection is actually wrong. read the letter →

arxiv 2506.12599 v3 pith:4AD6KCPY submitted 2025-06-14 eess.SP

Anthony Joseph Perre , Parker Huggins , Alphan Sahin This is my paper · ORCID
classification eess.SP
keywords BMOCZSBMOCZcarrierfrequencyoffsetzeroconstellationHuffmansequencesnon-coherentcommunicationDiZeTdecodercyclicpermutablecodes
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 variation of Huffman BMOCZ, called SBMOCZ, in which the angular separation between adjacent zeros is shrunk except at the first and last zero, squeezing the zero constellation into a shape with a distinct gap. The gap rotates when a carrier frequency offset (CFO) rotates the received polynomial, and the paper claims the receiver can find that rotation by locating the gap with a single N-point DFT and then correct it before decoding, with no channel coding. The reported error-rate results show that uncoded SBMOCZ remains functional under a CFO drawn uniformly over $[0,2\pi)$, at a cost of 1.46 dB in AWGN and 2.92 dB in fading relative to Huffman BMOCZ without a CFO. For coded operation, SBMOCZ with a (127,106)-BCH code is reported to gain 4 dB in BER over Huffman BMOCZ protected by a cyclically permutable code in a fading channel. The significance is that CFO correction no longer forces the use of cyclic codes, so standard error-correction codes become available under CFO.

What carries the argument

The central object is the SBMOCZ zero constellation with phase mapping $\varphi_k = \frac{(2\pi-\zeta)k}{K} + \frac{2\pi+\zeta(K-1)}{2K}$ and radius $r_{\mathrm{sb}} = \sqrt{1+2\lambda\sin\!\bigl(\frac{2\pi-\zeta}{2K}\bigr)}$, which introduces a gap in the zero constellation while preserving the conjugate-reciprocal zero structure. The argument is carried by two properties: Corollary 1, which asserts that the unit-circle magnitude profile $|X(e^{-j\theta})|^2$ is identical for every transmitted SBMOCZ message (so the gap's peak location does not depend on the data), and Lemma 1, which asserts the profile is even, so the peak is uniquely located. Together they imply that the peak of $|\tilde{Y}(e^{-j\theta})|$ shifts by exactly the CFO rotation, giving the estimator $\hat{\psi}\approx \frac{2\pi}{N}\arg\max_{n\in[N]}|\tilde{Y}(e^{-j2\pi n/N})|$, computable by a single $N$-point DFT. The corrected polynomial is then fed to the DiZeT decoder, which compares the received polynomial at conjugate-reciprocal zero pairs.

What would settle it

With $K=32$ and $\zeta=1/20$, compute the unit-circle magnitude $|X(e^{-j\theta})|^2$ for two messages, one with all zeros at the outer radius and one with alternating inner and outer radii; if the argmax of the two curves differs by more than the DFT resolution $2\pi/N$, the message-independence premise fails and the CFO estimate would carry a data-dependent bias.

Watch

Extended reading notes

Core claim

The paper's central claim is that a smooshed zero constellation makes CFO-induced rotation directly observable, so a receiver can estimate and remove it without any channel coding. The constellation is defined by a smooshing factor $\zeta$ that compresses the phases of adjacent zeros, leaving a gap in the zero constellation; under a CFO rotation $\psi$, the received polynomial becomes $\tilde{Y}(z)=Y(z e^{j\psi})$, and the gap shifts by $\psi$. The paper argues that $|X(e^{-j\theta})|^2$ on the unit circle is the same for every SBMOCZ message and has a unique maximum at the gap, so the CFO can be estimated as $\hat{\psi} = \arg\max_{\theta\in[0,2\pi)} |\tilde{Y}(e^{-j\theta})|$, implemented by one $N$-point DFT. After rotating the received polynomial back by $\hat{\psi}$, the standard DiZeT decoder recovers the bits. The authors demonstrate that this works for uncoded SBMOCZ across the full CFO range and that coded SBMOCZ with a BCH code outperforms Huffman BMOCZ with a cyclic permutable code by 4 dB in BER in a fading channel.

Load-bearing premise

The method assumes that every SBMOCZ message produces the same unit-circle magnitude profile, so the gap's peak location depends only on the CFO and not on the transmitted bits; the paper proves this only for the equispaced Huffman case and for the all-outer-radius SBMOCZ message, not for general SBMOCZ zero patterns with $\zeta>0$.

Editorial extensions

If this is right

  • Uncoded SBMOCZ operates under a CFO drawn uniformly from $[0,2\pi)$, where uncoded Huffman BMOCZ fails completely, accepting a 1.46 dB BER loss in AWGN and 2.92 dB in fading relative to Huffman BMOCZ without CFO.
  • The CFO correction requires only one $N$-point DFT, so it adds $O(N\log N)$ complexity instead of the overhead of cyclic-code construction and decoding.
  • Because CFO correction no longer depends on cyclic code structure, SBMOCZ can be paired with standard codes; the paper shows a (127,106)-BCH code gaining 4 dB in BER over ACPC-coded Huffman BMOCZ in fading.
  • Without a CFO, SBMOCZ is only modestly worse than Huffman BMOCZ in BER (about 1.3 dB in AWGN and 0.85 dB in fading) and in BLER (about 1.5 dB in AWGN and 1 dB in fading).
  • The smooshing factor $\zeta$ balances CFO-estimation reliability against noise-induced zero displacement, and the paper selects it by a parameter sweep over BER under CFO in AWGN.

Reading between the lines

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

  • Editorial inference: The gap-location strategy is not inherently tied to CFO; a constellation with a known non-uniform angular density could also make time offsets observable, though the paper does not analyze this case.
  • Editorial inference: Since the estimate resolution is limited by the DFT length $N$, a coarse-to-fine search around the peak could reduce residual CFO at low $\mathrm{E_b/N_0}$ without changing the constellation or adding channel coding.
  • Editorial inference: The message-independence of $|X(e^{-j\theta})|^2$ for $\zeta>0$ is testable offline for small $K$; if it fails for arbitrary messages, the CFO estimate would require either a training message or a modified constellation, which would be a natural next test of the paper's key premise.
  • Editorial inference: The 4 dB fading-channel BER gain over ACPC could stem from the BCH code's higher error-correction capability (three bits versus two for the ACPC) rather than from the CFO estimator itself; isolating these contributions would clarify where the gain comes from.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes a new BMOCZ zero constellation, SBMOCZ, in which the angular separation between consecutive zeros is reduced for all but one adjacent pair, creating a single larger gap in the zero constellation. The receiver estimates CFO by evaluating the received polynomial on the unit circle and taking the argument that maximizes the magnitude, implemented via an N-point DFT, then applies the inverse rotation before DiZeT decoding. Simulations for K=128 uncoded transmission in AWGN and flat fading show that SBMOCZ operates under a full-range CFO with a 1.46 dB loss in AWGN and a 2.92 dB loss in fading relative to Huffman BMOCZ without CFO, while uncoded Huffman BMOCZ fails. For K=127 coded transmission, SBMOCZ with a (127,106)-BCH code is reported to achieve a 4 dB BER gain over a Huffman BMOCZ baseline using a (127,106)-ACPC in fading, with comparable performance in AWGN.

Significance. If the key assumption of a unique, message-independent unit-circle peak can be rigorously established for the operating parameters, the paper offers a simple and low-complexity CFO-correction mechanism for BMOCZ that avoids the code-structure restrictions of the ACPC approach. The message independence of the SBMOCZ unit-circle profile is a useful property that follows from reciprocal-radius normalization, and the DFT implementation is a practical strength. The reported 4 dB fading gain over ACPC is an interesting system-level result. However, the central open issue is the lack of proof that the unit-circle profile has a unique global maximum with sufficient margin at the chosen smooshing factor; the current justification is heuristic, and the numerical results do not isolate the estimator's failure mechanism.

major comments (3)
  1. [Section III-B, Eq. (12)] The CFO estimator requires that |X(e^{-jθ})|² has a unique global maximum at θ=0 for the constellation used in the simulations (K=128, ζ=0.0117, r_sb≈1.0122). The paper supports this only with the zero-density heuristic and Fig. 2, which uses K=32 with ζ=3/100 and ζ=1/20. For the actual operating point, the profile is a small perturbation of the equispaced Huffman constellation, whose unit-circle profile has K degenerate maxima, and the largest gap is only about 24% larger than the nominal inter-zero spacing. The paper should either prove the uniqueness and quantify the peak-to-secondary-maximum margin as a function of ζ and K, or provide a numerical profile for the exact operating parameters. Without this, the estimator (12) is not guaranteed to lock onto the intended gap, and the low-SNR behavior in Fig. 3(a) may be dominated by secondary-maximum mislocks.
  2. [Section III-B, Corollary 1 and Appendix B] Corollary 1 is justified by a citation to [4], but [4] establishes the constant-autocorrelation property in the Huffman/equispaced setting. Since SBMOCZ deliberately changes the angle mapping, the message independence needed for (12) must be proven for the angle set in (8). The claim is in fact true for any fixed angle set with conjugate-reciprocal radii: using |x_K|²=(K+1)/∏(1+|α_k|²), the factor |e^{-jθ}-r_k e^{jφ_k}|²/(1+r_k²) is equal to the corresponding reciprocal-radius factor. This identity is not shown. Appendix B proves Lemma 1 only for the all-outer-radius message and then invokes Corollary 1; a direct proof of Corollary 1 would make the argument complete.
  3. [Sections IV-A and IV-B] The smooshing factor ζ is selected by a parameter sweep that minimizes BER under CFO in AWGN, and the same CFO AWGN scenario is then used to report the uncoded BER curves in Fig. 3. This is in-sample tuning: the reported losses (1.46 dB in AWGN, 2.92 dB in fading) are conditional on the tuned ζ and may be optimistic. Please provide a separate validation scenario, a sensitivity analysis over ζ, or a principled selection rule that does not depend on the exact CFO distribution used in the evaluation.
minor comments (4)
  1. [Appendix A] There is a typo: 'radial seperation' should be 'radial separation'.
  2. [Section III-B] The notation switches between |Y(z)|, |Y(e^{-jθ})|, and |\tilde Y(e^{-jθ})| without always making clear whether the magnitude is evaluated on the unit circle; please define the convention once and use it consistently.
  3. [Section III-A] The sentence 'SBMOCZ with ζ=0 and Huffman BMOCZ interchangeably' should be checked: Eq. (8) for ζ=0 yields a global rotation by π/K relative to the standard Huffman phase mapping φ_k=2πk/K. The equivalence presumably holds because DiZeT decoding is invariant to a common rotation, but this should be stated explicitly.
  4. [Fig. 2] Please label the axes of the unit-circle magnitude plots and state whether the vertical axis is |X(e^{-jθ})| or its square; the caption says |X(z)|² but the text also refers to |X(z)|.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the CFO estimator is a modulation-design feature, not a fitted prediction, and the message-independence of the unit-circle magnitude profile is an independently checkable algebraic identity.

full rationale

The paper's derivation chain is self-contained in the relevant sense. The CFO estimator (12) is based on a known geometric property of the proposed constellation: the zero gap rotates under CFO, and the unit-circle magnitude profile is message-independent (Corollary 1). Although Corollary 1 is attributed to the external reference [4], the underlying identity follows directly from the reciprocal-radius normalization and the coefficient norm constraint for any fixed angle set, so the argument does not reduce to the paper's own claims. Lemma 1's symmetry proof is algebraic and correct for the phase mapping in (8). The selection of the smooshing factor ζ by a parameter sweep on the CFO-in-AWGN BER is in-sample tuning, but the reported BER curves are simulations, not predictions derived from the fitted value; no fitted parameter is renamed as a predicted result. The unproved uniqueness and dominance of the unit-circle peak for the chosen parameters (K=128, ζ=0.0117) is an unverified assumption and a robustness risk, not a circular step. Self-citations ([5], [7]) appear only in contextual or supporting roles and are not load-bearing for the CFO-estimation claim. Overall, no prediction or central result reduces to its own inputs.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central design relies on the Huffman message-independence of the autocorrelation, imported from [4] via Corollary 1, without re-deriving it for the modified angles; this is the main unproven input. The CFO model and flat-fading assumption are standard domain assumptions. Parameter ζ is tuned to the test scenario.

free parameters (3)
  • smooshing factor ζ = 0.0117 (K=128), 0.0130 (K=127)
    Selected by parameter sweep to minimize BER under CFO in AWGN (Section IV). It sets the gap size and the radius per (9); small ζ gives weak CFO correction, large ζ increases zero perturbation under noise.
  • trade-off factor λ = 0.5
    Inherited from Huffman BMOCZ design [4]; balances radial vs angular zero separation. Fixed at 0.5 in all simulations.
  • DFT size for CFO search N = 1024
    Resolution of the angular search in (13); a computational choice, not fitted to channel data.
assumptions (4)
  • standard math Fundamental theorem of algebra and DTFT/autocorrelation relations (Proposition 1, Section III-B).
    Used to relate unit-circle evaluation of the received polynomial to the autocorrelation of the transmitted coefficients.
  • domain assumption Flat-fading channel with L=1 (Section II-A).
    Ensures the received polynomial has the same number of zeros as the transmitted polynomial; motivated by OFDM mapping from [5].
  • domain assumption CFO rotates zeros by ψ modulo 2π per (6), with ψ uniform on [0,2π) in simulations.
    Footnote 1 acknowledges the full-range rotation is unrealistic, but it is used to align with [6].
  • ad hoc to paper Autocorrelation of the transmitted BMOCZ sequence is independent of the message bits (Corollary 1).
    Imported from [4] and applied to SBMOCZ without proof; generally false when the zero angles are not equispaced by 2π/K, and Appendix B only proves the symmetric all-outer case.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Smooshed BMOCZ Zero Constellation for CFO Estimation Without Channel Coding." pith.science (2026). https://pith.science/paper/4AD6KCPY

@misc{pith2026250612599,
  author       = {Pith},
  title        = {Pith review of: A Smooshed BMOCZ Zero Constellation for CFO Estimation Without Channel Coding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4AD6KCPY}},
  note         = {Machine review of arXiv:2506.12599}
}
read the original abstract

In this study, we propose a new binary modulation on conjugate-reciprocal zeros (BMOCZ) zero constellation, which we call smooshed binary modulation on conjugate-reciprocal zeros (SBMOCZ), to address carrier frequency offset (CFO)-induced zero rotation without depending on channel coding. In our approach, we modify the phase mapping of Huffman BMOCZ by shrinking the angle between adjacent zeros, except for the first and last, to introduce a gap in the zero constellation. By discerning the gap location in the received polynomial, the receiver can estimate and correct the phase rotation. We demonstrate the error rate performance of SBMOCZ relative to Huffman BMOCZ, showing that SBMOCZ addresses a CFO-induced rotation at the cost of a modest performance reduction compared to Huffman BMOCZ in the absence of a CFO. Finally, we compare SBMOCZ to Huffman BMOCZ using a cyclically permutable code (CPC), showing a 4 dB bit error rate (BER) improvement in a fading channel, while demonstrating comparable performance across other simulations.

Figures

Figures reproduced from arXiv: 2506.12599 by the authors.

Figure 1
Figure 1. Comparison of Huffman BMOCZ and SBMOCZ for [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. |X(z)| 2 evaluated on the unit circle for K = 32. magnitude of X(e−jθ ) for θ ∈ [0, 2π) corresponds to the discrete-time Fourier transform (DTFT) of the auto-correlation sequence a = [a−K, a−K+1, . . . , aK] ∈ C 2K+1 for x. In particular, we have [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Comparison of uncoded schemes for K = 128. The SBMOCZ scheme is configured with rsb = 1.0122 and ζ = 0.0117, while Huffman BMOCZ uses rhb = 1.0122. and fading channels. Without a CFO, SBMOCZ performs roughly 1.3 dB worse than Huffman BMOCZ in AWGN and 0.85 dB worse in fading. However, when a CFO is introduced, Huffman BMOCZ fails, as indicated by the red curve, while SBMOCZ remains functional but loses 1.46 dB in AW… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

10 extracted references · 9 canonical work pages

  1. [4]

    MOCZ for blind short-pa cket communication: Basic principles,

    P . Walk, P . Jung, and B. Hassibi, “MOCZ for blind short-pa cket communication: Basic principles,” IEEE Transactions on Wireless Com- munications, vol. 18, no. 11, pp. 5080–5097, 2019

    work page 2019

  2. [1]

    Non-coherent and backscatter communications: Enab ling ultra- massive connectivity in 6G wireless networks,

    S. J. Nawaz, S. K. Sharma, B. Mansoor, M. N. Patwary, and N. M. Khan, “Non-coherent and backscatter communications: Enab ling ultra- massive connectivity in 6G wireless networks,” IEEE Access, vol. 9, pp. 38 144–38 186, 2021

    work page 2021

  3. [2]

    Noncoherent ultra-wideband systems,

    K. Witrisal, G. Leus, G. J. Janssen, M. Pausini, F. Troesc h, T. Zasowski, and J. Romme, “Noncoherent ultra-wideband systems,” IEEE Signal Processing Magazine, vol. 26, no. 4, pp. 48–66, 2009

    work page 2009

  4. [3]

    Sixty years of coherent versus non -coherent tradeoffs and the road from 5G to wireless futures,

    C. Xu, N. Ishikawa, R. Rajashekar, S. Sugiura, R. G. Maund er, Z. Wang, L.-L. Y ang, and L. Hanzo, “Sixty years of coherent versus non -coherent tradeoffs and the road from 5G to wireless futures,” IEEE Access, vol. 7, pp. 178 246–178 299, 2019

    work page 2019

  5. [5]

    On the optimal radius and subca rrier mapping for binary modulation on conjugate-reciprocal zeros,

    P . Huggins and A. S ¸ ahin, “On the optimal radius and subca rrier mapping for binary modulation on conjugate-reciprocal zeros,” in Proc. IEEE Military Communications Conference (MILCOM) , 2024, pp. 1–6

    work page 2024

  6. [6]

    MOCZ for blind short-packet communication: Practical aspects,

    P . Walk, P . Jung, B. Hassibi, and H. Jafarkhani, “MOCZ for blind short-packet communication: Practical aspects,” IEEE Transactions on Wireless Communications, vol. 19, no. 10, pp. 6675–6692, 2020

    work page 2020

  7. [7]

    Over-the-air majority vote computation wit h modulation on conjugate-reciprocal zeros,

    A. S ¸ ahin, “Over-the-air majority vote computation wit h modulation on conjugate-reciprocal zeros,” IEEE Transactions on Wireless Communi- cations, vol. 23, no. 11, pp. 17 714–17 726, 2024

    work page 2024

  8. [8]

    Integrated sensing and communication with MOCZ w ave- form,

    S. K. Dehkordi, P . Jung, P . Walk, D. Wieruch, K. Heuermann , and G. Caire, “Integrated sensing and communication with MOCZ w ave- form,” arXiv preprint arXiv:2307.01760 , 2023

    arXiv 2023

Show all 10 references
  1. [9]

    The design of Huffman sequences,

    M. H. Ackroyd, “The design of Huffman sequences,” IEEE Transactions on Aerospace and Electronic Systems , no. 6, pp. 790–796, 1970

    1970

  2. [10]

    Short-message commun ication and FIR system identification using Huffman sequences,

    P . Walk, P . Jung, and B. Hassibi, “Short-message commun ication and FIR system identification using Huffman sequences,” in Proc. IEEE International Symposium on Information Theory (ISIT) , 2017, pp. 968– 972

    2017

Pith tools

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