Nonexistence Results and Constructions for Signed Difference Sets

Jiayan Wu; Zhiwen He

REVIEW 2 major objections 4 minor 14 references

Nonexistence Results and Constructions for Signed Difference Sets

T0 review · 2 major / 4 minor · reviewed 2026-08-28 · deepseek-v4-flash

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

Pith's one-line read Four new nonexistence obstructions join two new existence constructions to settle 45,361 open database cases for signed difference sets in abelian groups up to order 499.

desk verdict Solid nonexistence criteria, a nice PCP construction, and a (125,28,3) existence proof that is internally inconsistent as printed. read the letter →

arxiv 2608.22936 v1 pith:JWZYZ52L submitted 2026-08-24 math.CO

Zhiwen He , Jiayan Wu This is my paper · ORCID

classification math.CO

keywords differencesetscasesconstructionscriteriaexistencemethodnonexistence

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

A signed difference set is a way of writing a group element with coefficients -1, 0, or +1 so that all nonzero shifts of its correlation are equal. Such sets generalize ordinary difference sets and are related to weighing matrices and ternary sequences with flat autocorrelation. The paper studies them in finite abelian groups.

The authors prove four conditions that make many signed difference sets impossible. The first uses prime ideals in cyclotomic fields; the other three count how many +1 and -1 entries lie in each coset of a subgroup and compare parities. When these conditions are run on a published database of 67,823 open cases with group sizes from 9 to 499, they rule out 45,361 parameter-group combinations.

The paper also constructs new examples. One construction takes partial difference sets from Desarguesian spreads to make an infinite family in elementary abelian 2-groups. The other uses quartic multiplicative characters to build an explicit signed difference set in the group C_5^3, with parameters (125,28,3). The headline numbers depend on the external database, and the written proof of the 125-case contains a repairable error in a Gauss sum identity; the construction itself appears to work once the identity is corrected.

Extended reading notes

Core claim

The central assertion is that the four obstructions, applied cumulatively to the 67,823 open group-specific cases in the database with 9 <= v <= 499, rule out 45,361 cases, while the two existence constructions settle three further cases. If this is correct, the nonexistence census is advanced by 45,361 parameter-group pairs and the (125,28,3)-SDS in C_5^3 exists.

Load-bearing premise

The headline counts rest on the external La Jolla Combinatorics Repository (ref. [8]) and the authors' verification code (ref. [12]); the preprint does not reproduce the database snapshot or the exclusion log. If any of those open cases are misclassified in the database, or if the code is incomplete, the numbers 67,823, 45,361 and 'three further cases' would change. The mathematical theorems themselves do not depend on the database.

Share X Bluesky LinkedIn Reddit HN

Formalized claims in Lean

Claim #1: The central assertion is that the four obstructions, applied cumulatively to the 67,823 open group-specific cases in the database with 9 <= v <= 499, rule out 45,361 cases, while the two existence constructions settle three further cases. If this is correct, the nonexistence census is advanced by 45,361 parameter-group pairs and the (125,28,3)-SDS in C_5^3 exists.

/-- @claim 1 The central assertion is that the four obstructions, applied cumulatively to the 67,823 open group-specific cases in the database with 9 <= v <= 499, rule out 45,361 cases, while the two existence constructions settle three further cases. If this is correct, the nonexistence census is advanced by 45,361 parameter-group pairs and the (125,28,3)-SDS in C_5^3 exists. -/
def central_claim : Prop :=

Signed reviews

No signed human review yet.

Request a human review

A listed scientist reviews the paper for a fee and the review publishes here regardless of verdict. See the reviewers or get listed.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Two things to know. The nonexistence work is solid: the self-conjugate prime criterion is a signed adaptation of Turyn, and the support-refined quotient method with the C2, near-full-support, and small-defect parity obstructions is genuinely new and cleanly proved. The PCP-derived infinite family in elementary abelian 2-groups is explicit and checks out. The (125,28,3)-SDS construction has a real gap: the proof of Theorem 5.6 is internally inconsistent, so the existence half of the headline is not established as written.

The new material is worth credit. The support-refined quotient lemma is a nice way to expose parity information that ordinary intersection numbers miss. The three parity obstructions are simple, correct, and do real census work. The PCP construction from Desarguesian spreads gives an infinite family and the two existence cases at v=64 and 256 look correct. The authors also made the right move from the circularity viewpoint: all obstructions derive from the defining equation, no fitting to the target.

The soft spots. Theorem 5.6 fails as printed. Lemma 5.5 states J(σ,σ)=-1+2i but then uses the conjugate -1-2i in the product J(ρ,ρ)J(σ,σ) to get 5-10i; with the stated values the product is -11+2i. Later the proof prints C=-A, but the τ=1 evaluation uses C=-conj(A). The two errors cancel in the right direction—the construction may well be valid after repair—but the character verification in the manuscript is not a valid derivation. The census counts (67,823, 45,361, etc.) rest on an external database snapshot and verification code that are not reproduced in the preprint; the theorems do not depend on that data, but the headline numbers do. That is a deposit issue, not a mathematical one.

Who this is for: specialists in difference sets and sequence design. The paper advances the SDS census substantially and gives tools others will reuse. It deserves a serious referee. My recommendation: engage with it, but require a corrected proof of Theorem 5.6 (or a machine-checked/independent verification) and, at revision, a reproducible log or checksum for the census counts.

Referee Report

2 major / 4 minor

Summary. The paper studies signed difference sets (SDSs) in finite abelian groups. It establishes four nonexistence obstructions: a self-conjugate prime character-norm obstruction (Theorem 3.2), a C2-quotient perfect-square obstruction (Theorem 4.2), a near-full-support parity obstruction (Theorem 4.4), and a small-defect parity obstruction (Theorem 4.7). It also gives two existence constructions: an infinite family in elementary abelian 2-groups from PCP-type regular partial difference sets arising from Desarguesian spreads (Theorem 5.3), and an explicit (125,28,3)-SDS in C5^3 built from quartic multiplicative characters (Theorem 5.6). Cumulatively, the numerical census claims to rule out 45,361 open group-specific cases and to settle three further existence cases.

Significance. If the proofs and census are validated, the paper makes a substantial contribution to the SDS catalog: the nonexistence criteria are derived transparently from the defining group-ring identity without fitted parameters, the spread construction is explicit and gives an infinite family, and the (125,28,3) construction would settle a previously open existence case. The main theorems are clean and checkable, with the important exception of the character calculation in Theorem 5.6. The numerical impact relies on an external database and verification code, so the census claims are not independently reproducible from the manuscript alone.

major comments (2)

[§5.2, Lemma 5.5 and Theorem 5.6] The printed proof of Theorem 5.6 is internally inconsistent. Lemma 5.5 states J(σ,σ)=-1+2i and then asserts J(ρ,ρ)J(σ,σ)=5-10i, but the product of J(ρ,ρ)=3+4i and J(σ,σ)=-1+2i is -11+2i; the value 5-10i is actually J(ρ,ρ) times the conjugate of J(σ,σ). In the Gauss-sum reduction, the proof then asserts C=G25(ρ^{-1})G5(σ)=-A. The standard identity G_q(χ^{-1})=χ(-1)\overline{G_q(χ)}, together with ρ(-1)=1 and σ(-1)=-1, gives C=-\overline{A}, not C=-A. The later evaluation for τ=1, namely 4Θ(D)=5+B+2iY, is precisely what follows from C=-\overline{A} when A=X+iY, so the derivation mixes two mutually inconsistent substitutions. Since Theorem 5.6 is one of the two existence constructions and is used to claim one of the three settled open cases, the existence claim is not established as printed. The calculation appears repairable by replacing C=-A with C=-\overline{A} and correcting the product in Lemma 5.5, but the proof must be corrected before acceptance.
[§1, Table 1, Examples 3.4, 4.3, 4.6, 4.8] The nonexistence census counts (44,872; 456; 29; 4; total 45,361) depend on the external La Jolla repository [8] and the verification code [12], but the manuscript does not include the database snapshot, the exclusion log, or the code output. As a result, the headline numerical claims cannot be independently checked from the paper alone. Please provide the exact snapshot and a full log of excluded cases as supplementary material, or state precisely how to reproduce the counts from reference [12].

minor comments (4)

[§4.1, Theorem 4.2] The proof says 'applying Eq. (4.3) with y=0', but Eq. (4.3) is stated only for nonzero y; in G/H≅C2 the unique nonzero y gives the same coefficient, so the wording should be adjusted.
[§4, throughout] Expressions such as 'pm−1' and 'pm−1−d' are ambiguous; they should be typeset as p^{m-1} and p^{m-1}-d for readability.
[§5.2, proof of Theorem 5.6] The line 'Since G_q(χ)G_q(χ)=q' should read 'Since |G_q(χ)|^2=q'.
[§3, Example 3.4] The notation SDS(37,13,1,[37]) is used before the bracket notation for group type is defined; please introduce or explain this notation.

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

No free parameters or invented entities. The central claims rest on standard number-theoretic identities, a standard PDS parameter list, and the external database. The Gauss-sum identity error does not introduce a free parameter but is a proof defect.

assumptions (4)

standard math Standard cyclotomic field facts (prime ideal decomposition, Frobenius automorphism, complex conjugation).
Invoked in Theorem 3.2 and recalled in Section 2.2 with citations to Ireland-Rosen [13].
standard math Correct Gauss-Jacobi sum identities, including G(chi^-1) = chi(-1) conj(G(chi)) and G(chi)^2 = G(chi^2)J(chi,chi).
Used in Theorem 5.6. The paper misstates one of these identities; the corrected version is needed for the proof to go through.
domain assumption Parameters of ell-PCP-type regular partial difference sets from Desarguesian spreads (Proposition 5.2).
Taken from Ma [14] without proof and is the basis of Theorem 5.3.
domain assumption The external signed difference set database accurately lists open and closed group-specific cases.
Used for all census counts (67,823 open, 45,361 excluded, three settled). Not independently verified in this review.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Nonexistence Results and Constructions for Signed Difference Sets." pith.science (2026). https://pith.science/paper/JWZYZ52L

@misc{pith2026260822936,
  author       = {Pith},
  title        = {Pith review of: Nonexistence Results and Constructions for Signed Difference Sets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JWZYZ52L}},
  note         = {Machine review of arXiv:2608.22936}
}

abstract

Signed difference sets (SDSs) extend ordinary difference sets by allowing negative coefficients. We establish new nonexistence criteria and existence constructions for SDSs in finite abelian groups. Using the classical self-conjugate-prime method and a support-refined quotient method, we derive four obstructions, including the $C_2$-quotient, near-full-support, and small-defect obstructions. Applied cumulatively to the $67{,}823$ open group-specific cases in the database with $9\leq v\leq499$, these criteria rule out $45{,}361$ cases. For existence, we construct an infinite family from PCP-type regular partial difference sets arising from Desarguesian spreads and an explicit $(125,28,3)$-SDS in $C_5^3$ using quartic multiplicative characters. These constructions settle three further open cases.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

[8]

La jolla combinatorics repository,

D. M. Gordon, “La jolla combinatorics repository,” https://www.dmgordon.org, 2022, accessed 23 August 2026

work page 2022
[12]

SDS Open-Case Verification,

Z. He, “SDS Open-Case Verification,” https://github.com/zhiwenhe94/sds-open-case-verification, 2026, Python code and data for verifying the open-case counts. Accessed: Aug. 23, 2026

work page 2026
[1]

Spectrally-constrained sequences: bounds and constructions,

Z. Liu, Y . L. Guan, U. Parampalli, and S. Hu, “Spectrally-constrained sequences: bounds and constructions,”IEEE Trans. Inform. Theory, vol. 64, no. 4, pp. 2571–2582, 2018. [Online]. Available: https://doi.org/10.1109/TIT.2018.2800012

work page arXiv 2018
[2]

T. W. Cusick, C. Ding, and A. Renvall,Stream ciphers and number theory, revised ed., ser. North-Holland Mathe- matical Library. Elsevier Science B.V ., Amsterdam, 2004, vol. 66

work page 2004
[3]

Ding,Codes from difference sets

C. Ding,Codes from difference sets. World Scientific Publishing Co. Pte. Ltd., Hackensack, NJ, 2015

work page 2015
[4]

Synthesizing low autocorrelation and low PAPR OFDM sequences under spectral constraints through convex optimization and GS algorithm,

L.-S. Tsai, W.-H. Chung, and D.-s. Shiu, “Synthesizing low autocorrelation and low PAPR OFDM sequences under spectral constraints through convex optimization and GS algorithm,”IEEE Trans. Signal Process., vol. 59, no. 5, pp. 2234–2243, 2011. [Online]. Available: https://doi.org/10.1109/TSP.2011.2108652

work page arXiv 2011
[5]

Yang , author L

Z. Ye, Z. Zhou, Z. Liu, X. Tang, and P. Fan, “New spectrally constrained sequence sets with optimal periodic cross-correlation,”IEEE Trans. Inform. Theory, vol. 69, no. 1, pp. 610–625, 2023. [Online]. Available: https://doi.org/10.1109/tit.2022.3205380

work page arXiv 2023
[6]

Signed difference sets,

D. M. Gordon, “Signed difference sets,”Des. Codes Cryptogr., vol. 91, no. 5, pp. 2107–2115, 2023. [Online]. Available: https://doi.org/10.1007/s10623-022-01171-8

work page doi:10.1007/s10623-022-01171-8 2023

Show all 14 references

[7]

New constructions of signed difference sets,

Z. He, T. Chen, and G. Ge, “New constructions of signed difference sets,”Des. Codes Cryptogr., vol. 92, no. 8, pp. 2323–2340, 2024. [Online]. Available: https://doi.org/10.1007/s10623-024-01389-8

2024 doi
[9]

A survey of difference sets,

M. Hall, Jr., “A survey of difference sets,”Proc. Amer. Math. Soc., vol. 7, pp. 975–986, 1956. [Online]. Available: https://doi.org/10.2307/2033024

1956 doi
[10]

Character sums and difference sets,

R. J. Turyn, “Character sums and difference sets,”Pacific J. Math., vol. 15, pp. 319–346, 1965. [Online]. Available: http://projecteuclid.org/euclid.pjm/1102996015

1965
[11]

The nonexistence of seven difference sets,

R. Turyn, “The nonexistence of seven difference sets,”Illinois J. Math., vol. 9, pp. 590–594, 1965. [Online]. Available: http://projecteuclid.org/euclid.ijm/1256059300

1965
[13]

Ireland and M

K. Ireland and M. Rosen,A classical introduction to modern number theory, 2nd ed., ser. Graduate Texts in Mathematics. Springer-Verlag, New York, 1990, vol. 84. [Online]. Available: https://doi.org/10.1007/ 978-1-4757-2103-4

1990
[14]

A survey of partial difference sets,

S. L. Ma, “A survey of partial difference sets,”Des. Codes Cryptogr., vol. 4, no. 3, pp. 221–261, 1994. [Online]. Available: https://doi.org/10.1007/BF01388454 SouthChinaNormalUniversity, Guangzhou510631, China Email address:zhiwen he@zju.edu.cn SouthChinaUniversity ofTechnolo...

1994 doi

Pith tools

Reviewed August 28, 2026 · model on record in the stance chip above.

[1] [8]

La jolla combinatorics repository,

D. M. Gordon, “La jolla combinatorics repository,” https://www.dmgordon.org, 2022, accessed 23 August 2026

work page 2022

[2] [12]

SDS Open-Case Verification,

Z. He, “SDS Open-Case Verification,” https://github.com/zhiwenhe94/sds-open-case-verification, 2026, Python code and data for verifying the open-case counts. Accessed: Aug. 23, 2026

work page 2026

[3] [1]

Spectrally-constrained sequences: bounds and constructions,

Z. Liu, Y . L. Guan, U. Parampalli, and S. Hu, “Spectrally-constrained sequences: bounds and constructions,”IEEE Trans. Inform. Theory, vol. 64, no. 4, pp. 2571–2582, 2018. [Online]. Available: https://doi.org/10.1109/TIT.2018.2800012

work page arXiv 2018

[4] [2]

T. W. Cusick, C. Ding, and A. Renvall,Stream ciphers and number theory, revised ed., ser. North-Holland Mathe- matical Library. Elsevier Science B.V ., Amsterdam, 2004, vol. 66

work page 2004

[5] [3]

Ding,Codes from difference sets

C. Ding,Codes from difference sets. World Scientific Publishing Co. Pte. Ltd., Hackensack, NJ, 2015

work page 2015

[6] [4]

Synthesizing low autocorrelation and low PAPR OFDM sequences under spectral constraints through convex optimization and GS algorithm,

L.-S. Tsai, W.-H. Chung, and D.-s. Shiu, “Synthesizing low autocorrelation and low PAPR OFDM sequences under spectral constraints through convex optimization and GS algorithm,”IEEE Trans. Signal Process., vol. 59, no. 5, pp. 2234–2243, 2011. [Online]. Available: https://doi.org/10.1109/TSP.2011.2108652

work page arXiv 2011

[7] [5]

Yang , author L

Z. Ye, Z. Zhou, Z. Liu, X. Tang, and P. Fan, “New spectrally constrained sequence sets with optimal periodic cross-correlation,”IEEE Trans. Inform. Theory, vol. 69, no. 1, pp. 610–625, 2023. [Online]. Available: https://doi.org/10.1109/tit.2022.3205380

work page arXiv 2023

[8] [6]

Signed difference sets,

D. M. Gordon, “Signed difference sets,”Des. Codes Cryptogr., vol. 91, no. 5, pp. 2107–2115, 2023. [Online]. Available: https://doi.org/10.1007/s10623-022-01171-8

work page doi:10.1007/s10623-022-01171-8 2023