betaFeedback

WorksMathematics

Nonexistence of a quasi-symmetric 22-(33,15,35)(33,15,35) design

Abstract

It is shown that there exists no quasi-symmetric 22-(33,15,35)(33,15,35) design with intersection numbers 66 and 99. This is proved by considering the ternary code generated by the incidence vectors of the blocks and its dual code.

Introduction

A 22-(v,k,λ)(v,k,\lambda) design D\mathcal{D} is a pair consisting of a set P\mathcal{P} of vv points and a collection B\mathcal{B} of kk-element subsets of P\mathcal{P} (called blocks) such that every 22-element subset of P\mathcal{P} is contained in exactly λ\lambda blocks. The number of blocks that contain a given point is traditionally denoted by rr, and the total number of blocks is bb. We assume that 2<k<v2 < k < v throughout this paper. A design with no repeated block is called simple. In this paper, we consider only simple designs. Often a 22-(v,k,λ)(v,k,\lambda) design is simply called a 2-design. A 2-design can be represented by its (block by point) incidence matrix A=(aij)A = (a_{ij}), where aij=1a_{ij} = 1 if the jjth point is contained in the iith block and aij=0a_{ij} = 0 otherwise.

A 2-design is called quasi-symmetric if the number of points in the intersection of any two distinct blocks takes only two values xx and yy (x<yx < y). The values xx and yy are called the intersection numbers. Quasi-symmetric 2-designs are closely related to strongly regular graphs and codes (see, e.g., [2, 13, 14]). Four classes of quasi-symmetric 2-designs were given in [8], namely, multiples of symmetric 2-designs, Steiner systems with v>k2v > k^2, strongly resolvable 2-designs and residual designs of biplanes. A quasi-symmetric 2-design D\mathcal{D} is called exceptional if neither D\mathcal{D} nor its complementary design belongs to one of these four classes. It is a fundamental problem to determine whether there exists an exceptional quasi-symmetric 2-design for a given set of parameters. The feasible parameters of exceptional quasi-symmetric 22-(v,k,λ)(v,k,\lambda) designs with 2k≤v≤702k \le v \le70 are listed in [12 Table 48.25]. Codes have been used to investigate these existence problems (see, e.g., [3], [4], [6] and [7]). For recent nonexistence results based on restrictions on block graphs, see [9] and [10].

The existence of a quasi-symmetric 22-(33,15,35)(33,15,35) design is listed as an open question in [12 Table 48.25] (see also [2 Section 8.5.6]). Such a design necessarily has intersection numbers 6 and 9 (see [2 Section 8.5.6]). Among the parameter sets for exceptional quasi-symmetric designs listed as unresolved in that reference, this one has the smallest number of points. The aim of this paper is to establish the following theorem.

Theorem 1. There exists no quasi-symmetric 2-(33, 15, 35) design.

Remark 2. If such a design existed, its block graph, in which two distinct blocks are adjacent if and only if they intersect in exactly 9 points, would be a strongly regular graph with parameters (176, 45, 18, 9). Strongly regular graphs with these parameters are known to exist (see [1]).

Preliminaries

Our approach is to consider the ternary code generated by the incidence vectors of the blocks of a hypothetical design, together with its dual code. In this section, we introduce the notation and definitions used in the proof. Throughout the paper, the ring or field in which a calculation is performed is understood from the context. We specify it explicitly whenever necessary to avoid ambiguity.

Ternary codes

Let F3={0,1,2}\mathbb{F}_3 = \{0,1,2\} denote the finite field of order 3. A ternary code CC of length nn is a vector subspace of F3n\mathbb{F}_3^n. A vector of CC is called a codeword of $C.

The weight wt⁡(x)\operatorname{wt}(x) of a vector x∈F3nx \in\mathbb{F}_3^n is the number of nonzero components of xx.

The dual code C⊥C^\perp of a ternary code CC of length nn is defined as follows:

C⊥={x∈F3n∣x⋅y=0 for all y∈C},C^\perp= \{x \in\mathbb{F}_3^n \mid x \cdot y = 0\ \text{for all } y \in C\},

where

x⋅y=∑i=1nxiyix \cdot y = \sum_{i=1}^{n} x_i y_i

for x=(x1,x2,…,xn)x = (x_1,x_2,\ldots,x_n) and y=(y1,y2,…,yn)∈F3ny = (y_1,y_2,\ldots,y_n) \in\mathbb{F}_3^n. A ternary code CC is called self-orthogonal if C⊆C⊥C \subseteq C^\perp.

Maps and inner product

We employ the following maps and inner product. Define a map ϕ:F3→Z\phi: \mathbb{F}_3 \to\mathbb{Z} by

ϕ(0)=0, ϕ(1)=1 and ϕ(2)=−1.\phi(0) = 0, \ \phi(1) = 1\ \text{and}\ \phi(2) = -1.

Let ρ:Z→F3\rho: \mathbb{Z} \to\mathbb{F}_3 denote the reduction map modulo 3. Then

ρ(ϕ(a))=a\rho(\phi(a)) = a

for every element a∈F3a \in\mathbb{F}_3. We extend both maps coordinatewise to vectors and use the same symbols for these extensions. Thus, for u=(u1,u2,…,un)∈F3nu = (u_1,u_2,\ldots,u_n) \in\mathbb{F}_3^n and for z=(z1,z2,…,zn)∈Znz = (z_1,z_2,\ldots,z_n) \in\mathbb{Z}^n, we write

ϕ(u)=(ϕ(u1),ϕ(u2),…,ϕ(un))∈Zn\phi(u) = (\phi(u_1),\phi(u_2),\ldots,\phi(u_n)) \in\mathbb{Z}^n

and

ρ(z)=(ρ(z1),ρ(z2),…,ρ(zn))∈F3n.\rho(z) = (\rho(z_1),\rho(z_2),\ldots,\rho(z_n)) \in\mathbb{F}_3^n.

In particular, ρ(ϕ(u))=u\rho(\phi(u)) = u for every vector u∈F3nu \in\mathbb{F}_3^n. For vectors a=(a1,a2,…,an)a = (a_1,a_2,\ldots,a_n) and b=(b1,b2,…,bn)∈Znb = (b_1,b_2,\ldots,b_n) \in\mathbb{Z}^n, define their integer inner product by

⟨a,b⟩Z=∑i=1naibi.\langle a,b\rangle_{\mathbb{Z}} = \sum_{i=1}^{n} a_i b_i.

Proof of Theorem 1

The aim of this section is to give a proof of Theorem 1 that is as elementary as possible.

Ternary code associated with the design

Suppose that there exists a quasi-symmetric 2-(33, 15, 35) design D=(X,B)D=(X,\mathcal{B}) with intersection numbers 6 and 9. We may assume that X={1,2,…,33}X=\{1,2,\ldots,33\}. For each subset W⊆XW\subseteq X, let W⃗∈{0,1}33\vec{W}\in\{0,1\}^{33} denote its characteristic vector.

Let AA be the 176×33176\times33 incidence matrix of DD. Then the (B,p)(B,p)-entry of AA is 1 if p∈Bp\in B and 0 otherwise, and the row indexed by BB is B⃗\vec{B}. Since r=80r=80 and λ=35\lambda=35, we have

ATA=45I33+35J33and∑B∈BB⃗=801(1)A^{T}A=45I_{33}+35J_{33}\quad\text{and}\quad\sum_{B\in\mathcal{B}}\vec{B}=80\mathbf{1} \tag*{(1)}

over Z\mathbb{Z}, where ATA^{T} denotes the transpose of AA, I33I_{33} is the identity matrix of order 33, J33J_{33} is the 33×3333\times33 all-ones matrix and 1\mathbf{1} denotes the all-ones vector of length 33.

Let C(D)C(D) be the ternary code generated by the rows of AA, that is,

C(D)=⟨B⃗∣B∈B⟩⊆F333.C(D)=\langle\vec{B}\mid B\in\mathcal{B}\rangle\subseteq\mathbb{F}_{3}^{33}.

The 176×176176\times176 integer matrix AATAA^{T} has diagonal entries 15 and off-diagonal entries 6 or 9, since DD has intersection numbers 6 and 9. Since all these entries are divisible by 3, we have AAT≡O(mod3)AA^{T}\equiv O\pmod{3}, where OO is the 176×176176\times176 zero matrix. Thus, C(D)C(D) is self-orthogonal. Reducing the second equality in (1) modulo 3, we obtain

C(D)∋∑B∈BB⃗=21.C(D)\ni\sum_{B\in\mathcal{B}}\vec{B}=21.

It follows that

1∈C(D)⊆C(D)⊥.(2)\mathbf{1}\in C(D)\subseteq C(D)^\perp. \tag*{(2)}

Restrictions on (0, 1)-vectors in C(D)⊥C(D)^\perp

Let u=(u1,u2,…,u33)∈C(D)⊥u=(u_1,u_2,\ldots,u_{33})\in C(D)^\perp. For each block B∈BB\in\mathcal{B}, let

t(u)=∑p=133ϕ(up)∈ZandSB=∑p∈Bϕ(up)∈Z.t(u)=\sum_{p=1}^{33}\phi(u_p)\in\mathbb{Z}\quad\text{and}\quad S_B=\sum_{p\in B}\phi(u_p)\in\mathbb{Z}.

Lemma 3. Let uu be a vector of C(D)⊥C(\mathcal{D})^\perp. Then

SB∈3Z,∑B∈BSB=80t(u) and∑B∈BSB2=45wt⁡(u)+35t(u)2.(3)\begin{aligned} S_B &\in3\mathbb{Z},\\ \sum_{B\in\mathcal{B}} S_B &= 80t(u)\ \text{and}\\ \sum_{B\in\mathcal{B}} S_B^2 &= 45\operatorname{wt}(u)+35t(u)^2. \tag*{(3)} \end{aligned}

Proof. Since ρ\rho is additive and ρ(ϕ(a))=a\rho(\phi(a))=a for every element a∈F3a\in\mathbb{F}_3, we have

ρ(SB)=∑p∈Bup=∑p=133up(B⃗)p=u⋅B⃗=0\rho(S_B)=\sum_{p\in B}u_p=\sum_{p=1}^{33}u_p(\vec{B})_p=u\cdot\vec{B}=0

for each block B∈BB\in\mathcal{B}. Here, (B⃗)p(\vec{B})_p denotes the ppth coordinate of B⃗\vec{B}. The last equality follows from u∈C(D)⊥u\in C(\mathcal{D})^\perp and B⃗∈C(D)\vec{B}\in C(\mathcal{D}). Thus, SBS_B is divisible by 3 for every block B∈BB\in\mathcal{B}.

Since r=80r=80, we have

∑B∈BSB=∑p=133ϕ(up)∣{B∈B∣p∈B}∣=80t(u).\sum_{B\in\mathcal{B}}S_B=\sum_{p=1}^{33}\phi(u_p)\lvert\{B\in\mathcal{B}\mid p\in B\}\rvert=80t(u).

The entry of Aϕ(u)TA\phi(u)^T indexed by BB is SBS_B. From (1), we obtain

∑B∈BSB2=(Aϕ(u)T)T(Aϕ(u)T)=ϕ(u)ATAϕ(u)T=45∑p=133ϕ(up)2+35(∑p=133ϕ(up))2=45wt⁡(u)+35t(u)2.\begin{aligned} \sum_{B\in\mathcal{B}}S_B^2 &=(A\phi(u)^T)^T(A\phi(u)^T)\\ &=\phi(u)A^TA\phi(u)^T\\ &=45\sum_{p=1}^{33}\phi(u_p)^2+35\left(\sum_{p=1}^{33}\phi(u_p)\right)^2\\ &=45\operatorname{wt}(u)+35t(u)^2. \end{aligned}

The last equality follows from ϕ(0)=0\phi(0)=0 and ϕ(1)2=ϕ(2)2=1\phi(1)^2=\phi(2)^2=1. This completes the proof.

For use in the following lemma, note that

SB=∑p∈Bϕ(up)=⟨ϕ(u),ϕ(B⃗)⟩ZS_B=\sum_{p\in B}\phi(u_p)=\langle\phi(u),\phi(\vec{B})\rangle_{\mathbb{Z}}

for every vector u∈C(D)⊥u\in C(\mathcal{D})^\perp and B∈BB\in\mathcal{B}.

Lemma 4. Suppose that u∈C(D)⊥u \in C(D)^{\perp} and that ϕ(u)\phi(u) has 6 entries equal to 1, 3 entries equal to −1-1 and 24 entries equal to 0. Then

SB∈{0,3}S_B \in\{0,3\}

for every block B∈BB \in\mathcal{B}.

Proof. Since ϕ(u)\phi(u) has 6 entries equal to 1, 3 entries equal to −1-1 and all remaining entries equal to 0, and SB=⟨ϕ(u),ϕ(B→)⟩ZS_B = \langle\phi(u),\phi(\overrightarrow{B})\rangle_{\mathbb{Z}}, we have −3≤SB≤6-3 \leq S_B \leq6. By Lemma 3, we have SB∈3ZS_B \in3\mathbb{Z}. It follows that

SB∈{−3,0,3,6}.S_B \in\{-3,0,3,6\}.

Thus, for every block B∈BB \in\mathcal{B}, we have

SB(SB−3)≥0,S_B(S_B-3) \geq0,

with equality if and only if SB∈{0,3}S_B \in\{0,3\}. Moreover, since wt⁡(u)=9\operatorname{wt}(u)=9 and t(u)=3t(u)=3, Lemma 3 gives

∑B∈BSB=240\sum_{B\in\mathcal{B}} S_B = 240

and

∑B∈BSB2=45⋅9+35⋅32=720.\sum_{B\in\mathcal{B}} S_B^2 = 45\cdot9+35\cdot3^2=720.

Hence, we have

∑B∈BSB(SB−3)=720−3⋅240=0.\sum_{B\in\mathcal{B}} S_B(S_B-3)=720-3\cdot240=0.

Therefore, we conclude that SB∈{0,3}S_B \in\{0,3\} for every block B∈BB \in\mathcal{B}. □

We now apply Lemma 3 to characteristic vectors of subsets of XX. Suppose that W⊆XW \subseteq X satisfies W→∈C(D)⊥\overrightarrow{W} \in C(D)^{\perp}. Let u=W→u=\overrightarrow{W} and w=∣W∣w=|W|. The vector uu is a (0,1)(0,1)-vector of weight ww. Since ϕ(0)=0\phi(0)=0 and ϕ(1)=1\phi(1)=1, we have

t(u)=∑p=133ϕ(up)=w.t(u)=\sum_{p=1}^{33}\phi(u_p)=w.

Moreover, we have

SB=∑p∈Bϕ(up)=∣B∩W∣.S_B=\sum_{p\in B}\phi(u_p)=|B\cap W|.

for each block B∈BB \in\mathcal{B}. By Lemma 3, SB=∣B∩W∣S_B = |B \cap W| is divisible by 3. Since ∣B∣=15|B| = 15, we also have

0≤∣B∩W∣≤min⁡(w,15).0 \le|B \cap W| \le\min(w,15).

Let

Iw={i∈3Z∣0≤i≤min⁡(w,15)}.\mathcal{I}_w = \{i \in3\mathbb{Z} \mid0 \le i \le\min(w,15)\}.

Thus, ∣B∩W∣∈Iw|B \cap W| \in\mathcal{I}_w for every B∈BB \in\mathcal{B}. For each i∈Iwi \in\mathcal{I}_w, define

ni=∣{B∈B∣∣B∩W∣=i}∣.(4)n_i = \left|\{B \in\mathcal{B} \mid|B \cap W| = i\}\right|. \tag*{(4)}

Thus, nin_i is the number of blocks meeting WW in exactly ii points.

Since

ni=∣{B∈B∣∣B∩W∣=i}∣=∣{B∈B∣SB=i}∣,n_i = \left|\{B \in\mathcal{B} \mid|B \cap W| = i\}\right| = \left|\{B \in\mathcal{B} \mid S_B = i\}\right|,

each block is counted exactly once in ∑i∈Iwni\sum_{i \in\mathcal{I}_w} n_i. Thus, we have

∑i∈Iwni=∣B∣=176.(5)\sum_{i \in\mathcal{I}_w} n_i = |\mathcal{B}| = 176. \tag*{(5)}

For each i∈Iwi \in\mathcal{I}_w, the nin_i blocks with SB=iS_B = i contribute inii n_i to ∑B∈BSB\sum_{B \in\mathcal{B}} S_B and i2nii^2 n_i to ∑B∈BSB2\sum_{B \in\mathcal{B}} S_B^2. Hence, we obtain

∑B∈BSB=∑i∈Iwini and∑B∈BSB2=∑i∈Iw∑B∈BSB=iSB2=∑i∈Iwi2ni.(6)\begin{aligned} \sum_{B \in\mathcal{B}} S_B &= \sum_{i \in\mathcal{I}_w} i n_i \text{ and}\\ \sum_{B \in\mathcal{B}} S_B^2 &= \sum_{i \in\mathcal{I}_w} \sum_{\substack{B \in\mathcal{B}\\S_B=i}} S_B^2 = \sum_{i \in\mathcal{I}_w} i^2 n_i. \tag*{(6)} \end{aligned}

Using (5) and (6), and applying Lemma 3 with wt⁡(u)=t(u)=w\operatorname{wt}(u) = t(u) = w, we obtain

∑i∈Iwni=176,\sum_{i \in\mathcal{I}_w} n_i = 176,
∑i∈Iwini=80w and\sum_{i \in\mathcal{I}_w} i n_i = 80w \text{ and}
∑i∈Iwi2ni=45w+35w2.(7)\sum_{i \in\mathcal{I}_w} i^2 n_i = 45w + 35w^2. \tag*{(7)}

where all equalities are taken over Z\mathbb{Z}. The following lemma restricts the possible weights of (0,1)(0,1)-vectors in C(D)⊥\mathcal{C}(\mathcal{D})^\perp.

Lemma 5. Suppose that x∈C(D)⊥∩{0,1}33x \in C(D)^\perp\cap\{0,1\}^{33}. Then

wt⁡(x)∈{0,15,18,33}.(8)\operatorname{wt}(x) \in\{0,15,18,33\}. \tag*{(8)}

Proof. Since 1∈C(D)1 \in C(D) by (2) and x∈C(D)⊥x \in C(D)^\perp, we have

ρ(wt⁡(x))=x⋅1=0.\rho(\operatorname{wt}(x))=x\cdot1=0.

Thus, wt⁡(x)\operatorname{wt}(x) is divisible by 3. Since 1∈C(D)⊆C(D)⊥1 \in C(D) \subseteq C(D)^\perp by (2) and x∈C(D)⊥x \in C(D)^\perp, the linearity of C(D)⊥C(D)^\perp gives 1+2x∈C(D)⊥1+2x \in C(D)^\perp. Moreover, since xx is a (0,1)(0,1)-vector, we have

1+2x∈{0,1}33 and wt⁡(1+2x)=33−wt⁡(x).1+2x \in\{0,1\}^{33}\text{ and }\operatorname{wt}(1+2x)=33-\operatorname{wt}(x).

Therefore, it suffices to rule out weights wt⁡(x)=3\operatorname{wt}(x)=3, 6, 9 and 12. We consider these weights in the order wt⁡(x)=3\operatorname{wt}(x)=3, 9, 6 and 12, since the later cases use the exclusions established in the earlier ones.

(i) Suppose that wt⁡(x)=3\operatorname{wt}(x)=3. Take u=xu=x in Lemma 3. Then, since xx is a (0,1)(0,1)-vector, we have wt⁡(u)=t(u)=3\operatorname{wt}(u)=t(u)=3 and 0≤SB≤30\leq S_B\leq3 for every block B∈BB\in\mathcal{B}. Since SBS_B is divisible by 3, we have SB∈{0,3}S_B\in\{0,3\}. In either case, SB2=3SBS_B^2=3S_B.

On the other hand, (3) gives

∑B∈BSB=80⋅3=240 and\sum_{B\in\mathcal{B}} S_B=80\cdot3=240\text{ and}
∑B∈BSB2=45⋅3+35⋅32=450.\sum_{B\in\mathcal{B}} S_B^2=45\cdot3+35\cdot3^2=450.

This contradicts

∑B∈BSB2=∑B∈B3SB=3∑B∈BSB=720.\sum_{B\in\mathcal{B}} S_B^2=\sum_{B\in\mathcal{B}}3S_B=3\sum_{B\in\mathcal{B}}S_B=720.

(ii) Suppose that wt⁡(x)=9\operatorname{wt}(x)=9. Take u=xu=x in Lemma 3. Then wt⁡(u)=t(u)=9\operatorname{wt}(u)=t(u)=9 and 0≤SB≤90\leq S_B\leq9 for every block B∈BB\in\mathcal{B}. Since SBS_B is divisible by 3, we have

SB∈{0,3,6,9}.S_B\in\{0,3,6,9\}.

Then we have

(SB−3)(SB−6)≥0.(S_B-3)(S_B-6)\geq0.

On the other hand, (3) and b=176b = 176 give

∑B∈B(SB−3)(SB−6)=∑B∈BSB2−9∑B∈BSB+18b=(45⋅9+35⋅92)−9(80⋅9)+18⋅176=−72.\sum_{B\in\mathcal{B}}(S_B-3)(S_B-6)=\sum_{B\in\mathcal{B}}S_B^2-9\sum_{B\in\mathcal{B}}S_B+18b =\left(45\cdot9+35\cdot9^2\right)-9(80\cdot9)+18\cdot176 =-72.

This is a contradiction.

(iii) Suppose that wt⁡(x)=6\operatorname{wt}(x)=6. Write x=W⃗x=\vec{W}, where W⊆XW\subseteq X and ∣W∣=6|W|=6. Let n0n_0, n3n_3 and n6n_6 be as defined in (4). From (7), we have

n0+n3+n6=176,n_0+n_3+n_6=176,
3n3+6n6=4803n_3+6n_6=480

and

9n3+36n6=1530.9n_3+36n_6=1530.

This gives 18n6=9018n_6=90. Since n6>0n_6>0, there is a block B∈BB\in\mathcal{B} with ∣B∩W∣=6|B\cap W|=6. Hence, we have W⊆BW\subseteq B. By the definition of C(D)C(\mathcal{D}) and (2), we have B⃗∈C(D)⊆C(D)⊥\vec{B}\in C(\mathcal{D})\subseteq C(\mathcal{D})^\perp. Since W⃗=x∈C(D)⊥\vec{W}=x\in C(\mathcal{D})^\perp by the assumption, we have

B⃗+2W⃗=B∖W→∈C(D)⊥∩{0,1}33\vec{B}+2\vec{W}=\overrightarrow{B\setminus W}\in C(\mathcal{D})^\perp\cap\{0,1\}^{33}

and

wt⁡(B∖W→)=15−6=9.\operatorname{wt}(\overrightarrow{B\setminus W})=15-6=9.

This contradicts the exclusion of weight 9 established above.

(iv) Suppose that wt⁡(x)=12\operatorname{wt}(x)=12. Write x=W⃗x=\vec{W}, where W⊆XW\subseteq X and ∣W∣=12|W|=12. Let n0n_0, n3n_3, n6n_6, n9n_9 and n12n_{12} be as defined in (4). From (7), we have

18(n0+n9+3n12)=∑i∈I12(i−3)(i−6)ni=(45⋅12+35⋅122)−9(80⋅12)+18⋅176=108.18(n_0+n_9+3n_{12})=\sum_{i\in I_{12}}(i-3)(i-6)n_i =\left(45\cdot12+35\cdot12^2\right)-9(80\cdot12)+18\cdot176 =108.

Thus, we have

n0+n9+3n12=6.(9)n_0+n_9+3n_{12}=6. \tag*{(9)}

First suppose that n12>0n_{12} > 0. Then there is a block B∈BB \in\mathcal{B} with W⊆BW \subseteq B. As in the case wt⁡(x)=6\operatorname{wt}(x) = 6, we have B⃗∈C(D)⊆C(D)⊥\vec{B} \in C(D) \subseteq C(D)^\perp and W⃗∈C(D)⊥\vec{W} \in C(D)^\perp. It follows that

B⃗+2W⃗=B∖W→∈C(D)⊥∩{0,1}33\vec{B} + 2\vec{W} = \overrightarrow{B \setminus W} \in C(D)^\perp\cap\{0,1\}^{33}

and

wt⁡(B∖W→)=15−12=3.\operatorname{wt}(\overrightarrow{B \setminus W}) = 15 - 12 = 3.

This contradicts the exclusion of weight 3 established above. Hence, we have n12=0n_{12} = 0.

Now suppose that n9>0n_9 > 0. Choose a block B∈BB \in\mathcal{B} with ∣B∩W∣=9|B \cap W| = 9. Let

u=B⃗+2W⃗∈C(D)⊥.u = \vec{B} + 2\vec{W} \in C(D)^\perp.

Since ∣B∖W∣=6|B \setminus W| = 6 and ∣W∖B∣=3|W \setminus B| = 3, the vector ϕ(u)\phi(u) has 6 entries equal to 1, 3 entries equal to −1-1 and 24 entries equal to 0. By Lemma 4, we have

⟨ϕ(u),ϕ(B⃗)⟩Z∈{0,3}.\langle\phi(u), \phi(\vec{B})\rangle_{\mathbb{Z}} \in\{0,3\}.

On the other hand, since u=B⃗+2W⃗∈F333u = \vec{B} + 2\vec{W} \in\mathbb{F}_3^{33}, we have up=1u_p = 1 for p∈B∖Wp \in B \setminus W and up=1+2=0u_p = 1 + 2 = 0 for p∈B∩Wp \in B \cap W. Thus, we have ϕ(up)=1\phi(u_p) = 1 for p∈B∖Wp \in B \setminus W and ϕ(up)=0\phi(u_p) = 0 for p∈B∩Wp \in B \cap W. Hence, we obtain

⟨ϕ(u),ϕ(B⃗)⟩Z=∑p∈Bϕ(up)=∣B∖W∣=15−9=6,\langle\phi(u), \phi(\vec{B})\rangle_{\mathbb{Z}} = \sum_{p \in B} \phi(u_p) = |B \setminus W| = 15 - 9 = 6,

a contradiction. Therefore, we have n9=0n_9 = 0.

From (9), we have n0=6n_0 = 6. Thus, there is a block B∈BB \in\mathcal{B} with B∩W=∅B \cap W = \varnothing. As in the case wt⁡(x)=6\operatorname{wt}(x) = 6, we have B⃗,W⃗∈C(D)⊥\vec{B}, \vec{W} \in C(D)^\perp. From (2), we have 1∈C(D)⊥\mathbf{1} \in C(D)^\perp. It follows that

1+2B⃗+2W⃗=X∖(B∪W)→∈C(D)⊥∩{0,1}33\mathbf{1} + 2\vec{B} + 2\vec{W} = \overrightarrow{X \setminus(B \cup W)} \in C(D)^\perp\cap\{0,1\}^{33}

and

wt⁡(X∖(B∪W)→)=33−15−12=6.\operatorname{wt}(\overrightarrow{X \setminus(B \cup W)}) = 33 - 15 - 12 = 6.

This contradicts the exclusion of weight 6 established above.

This completes the proof.

Identity for a ternary code and its dual code

For a ternary code CC of length 33, define

S(C)=∑x∈C∩{0,1}33(−1)wt(x)(33−2wt(x)).S(C)=\sum_{x\in C\cap\{0,1\}^{33}}(-1)^{\mathrm{wt}(x)}(33-2\mathrm{wt}(x)).

The following identity can be derived from the MacWilliams identity for complete weight enumerators (see, e.g., [5 Theorem 2.3.2]) by specialization and differentiation, together with the assumptions 1∈C⊆C⊥1\in C\subseteq C^\perp. A direct proof of Lemma 6 using character sums is given in Appendix A.

Lemma 6. Let CC be a ternary code of length 33. Suppose that 1∈C⊆C⊥1\in C\subseteq C^\perp. Then

S(C⊥)=−316∣C∣S(C).(10)S(C^\perp)=-\frac{3^{16}}{|C|}S(C). \tag*{(10)}

Proof of Theorem 1

For a ternary code CC of length 33, define the following:

B(C)=C∩{0,1}33.B(C)=C\cap\{0,1\}^{33}.

We are ready to give a proof of Theorem 1.

Proof of Theorem 1. Suppose that there exists a quasi-symmetric 2-(33, 15, 35) design DD with intersection numbers 6 and 9. We apply the preceding lemmas to the ternary code C(D)C(D).

Let EE be either C(D)C(D) or C(D)⊥C(D)^\perp. From (2), we have C(D)⊆C(D)⊥C(D)\subseteq C(D)^\perp. Thus, we have E⊆C(D)⊥E\subseteq C(D)^\perp in either case. If x∈B(E)x\in B(E), then by Lemma 5 we have

wt(x)∈{0,15,18,33}.\mathrm{wt}(x)\in\{0,15,18,33\}.

Let 0\mathbf{0} denote the zero vector. From (2), we have 1∈E\mathbf{1}\in E. Thus, we have 0,1∈B(E)\mathbf{0},\mathbf{1}\in B(E). Hence, B(E)B(E) contains exactly one vector of weight 0 and exactly one vector of weight 33. Therefore, the total contribution of these two vectors to S(E)S(E) is

1⋅(−1)0(33−0)+1⋅(−1)33(33−66)=33+33=66.(11)1\cdot(-1)^0(33-0)+1\cdot(-1)^{33}(33-66)=33+33=66. \tag*{(11)}

For every vector x∈B(E)x\in B(E), we have 1+2x∈E\mathbf{1}+2x\in E. Since xx is a (0,1)(0,1)-vector, we have

1+2x∈B(E)andwt(1+2x)=33−wt(x).\mathbf{1}+2x\in B(E)\quad\text{and}\quad\mathrm{wt}(\mathbf{1}+2x)=33-\mathrm{wt}(x).

The map x↦1+2xx \mapsto1+2x is its own inverse, and hence is a bijection on B(E)\mathcal{B}(E). It maps the vectors of weight 15 to those of weight 18, and conversely. Define

a15(B(E))=∣{x∈B(E)∣wt⁡(x)=15}∣.a_{15}(\mathcal{B}(E))=\left|\{x\in\mathcal{B}(E)\mid\operatorname{wt}(x)=15\}\right|.

Therefore, B(E)\mathcal{B}(E) contains exactly a15(B(E))a_{15}(\mathcal{B}(E)) vectors of each of these two weights.

Using the definition of S(E)S(E) and (11), we obtain

S(E)=∑x∈B(E)(−1)wt⁡(x)(33−2wt⁡(x))=66+a15(B(E))(−1)15(33−30)+a15(B(E))(−1)18(33−36)=66−6a15(B(E)).(12)\begin{aligned} S(E)&=\sum_{x\in\mathcal{B}(E)}(-1)^{\operatorname{wt}(x)}(33-2\operatorname{wt}(x))\\ &=66+a_{15}(\mathcal{B}(E))(-1)^{15}(33-30)+a_{15}(\mathcal{B}(E))(-1)^{18}(33-36)\\ &=66-6a_{15}(\mathcal{B}(E)). \tag*{(12)} \end{aligned}

The characteristic vectors of the 176 distinct blocks of D\mathcal{D} are distinct codewords of weight 15 in C(D)C(\mathcal{D}) whose entries are all 0 or 1. Since C(D)⊆C(D)⊥C(\mathcal{D})\subseteq C(\mathcal{D})^\perp, they also belong to C(D)⊥C(\mathcal{D})^\perp. Consequently, we have

a15(B(C(D)))≥176anda15(B(C(D)⊥))≥176.a_{15}(\mathcal{B}(C(\mathcal{D})))\ge176\quad\text{and}\quad a_{15}(\mathcal{B}(C(\mathcal{D})^\perp))\ge176.

Applying (12) to each of the two codes, we obtain

S(C(D))≤66−6⋅176=−990<0andS(C(\mathcal{D}))\le66-6\cdot176=-990<0\quad\text{and}
S(C(D)⊥)≤66−6⋅176=−990<0.S(C(\mathcal{D})^\perp)\le66-6\cdot176=-990<0.

On the other hand, since (2) gives 1∈C(D)⊆C(D)⊥1\in C(\mathcal{D})\subseteq C(\mathcal{D})^\perp, by Lemma 6 we have

S(C(D)⊥)=−316∣C(D)∣S(C(D)).S(C(\mathcal{D})^\perp)=-\frac{3^{16}}{|C(\mathcal{D})|}S(C(\mathcal{D})).

Since ∣C(D)∣>0|C(\mathcal{D})|>0 and S(C(D))<0S(C(\mathcal{D}))<0, the right-hand side is positive, contradicting S(C(D)⊥)<0S(C(\mathcal{D})^\perp)<0. Therefore, no such design exists. This completes the proof of Theorem 1. □\square

Other parameter sets

Finally, we consider other parameter sets for exceptional quasi-symmetric designs.

Among those listed as unresolved in [2 Section 8.5.6], the two with the smallest numbers of points, excluding the case settled by Theorem 1, are those of a 2-(35,14,13)(35, 14, 13) design with intersection numbers 5 and 8 and a 2-(39,12,22)(39, 12, 22) design with intersection numbers 3 and 6.

The former is ruled out by a recent result of Shpectorov and Zhao [11]. Indeed, its block graph, in which two blocks are adjacent when they intersect in 8 points, would be strongly regular with parameters (85,14,3,2)(85, 14, 3, 2), contrary to their nonexistence theorem. Together with Theorem 1, this leaves the case of a quasi-symmetric 2-(39,12,22)(39, 12, 22) design with intersection numbers 3 and 6 as the smallest remaining case.

The ternary code approach used in this paper can also be considered for quasi-symmetric 2-(v,k,λ)(v, k, \lambda) designs with intersection numbers xx and yy satisfying

k≡x≡y≡0(mod3)andr≠0(mod3).k \equiv x \equiv y \equiv0 \pmod{3}\quad\text{and}\quad r \ne0 \pmod{3}.

A 2-(39,12,22)(39, 12, 22) design with intersection numbers 3 and 6 satisfies these conditions, where r=76r = 76. However, the present argument does not establish its nonexistence. Indeed, suppose that such a design D\mathcal{D} exists, and let C(D)C(\mathcal{D}) be its ternary code generated by the rows of the incidence matrix of D\mathcal{D}. Since k≡x≡y≡0(mod3)k \equiv x \equiv y \equiv0 \pmod{3}, the code C(D)C(\mathcal{D}) is self-orthogonal. Moreover, since r≠0(mod3)r \ne0 \pmod{3}, C(D)C(\mathcal{D}) contains the all-ones vector. Define

S39(C(D))=∑z∈C(D)∩{0,1}39(−1)wt⁡(z)(39−2wt⁡(z)),S_{39}(C(\mathcal{D})) = \sum_{z \in C(\mathcal{D}) \cap\{0,1\}^{39}} (-1)^{\operatorname{wt}(z)}(39 - 2\operatorname{wt}(z)),

and define S39(C(D)⊥)S_{39}(C(\mathcal{D})^\perp) similarly. The same calculation as in Appendix § 5 gives

S39(C(D)⊥)=(−1)(39+1)/23(39−1)/2∣C(D)∣S39(C(D))=319∣C(D)∣S39(C(D)).S_{39}(C(\mathcal{D})^\perp) = (-1)^{(39+1)/2}\frac{3^{(39-1)/2}}{|C(\mathcal{D})|}S_{39}(C(\mathcal{D})) = \frac{3^{19}}{|C(\mathcal{D})|}S_{39}(C(\mathcal{D})).

The coefficient is positive, so this identity does not force S39(C(D)⊥)S_{39}(C(\mathcal{D})^\perp) and S39(C(D))S_{39}(C(\mathcal{D})) to have opposite signs. Thus, the final sign argument used in our proof does not yield a contradiction. The present approach may also be useful in studying the nonexistence of a quasi-symmetric 2-(45,18,34)(45, 18, 34) design with intersection numbers 6 and 9.

Acknowledgments. This work was supported by JSPS KAKENHI Grant Numbers 23K25784 and 25K21995.

Use of generative AI. The author used OpenAI’s ChatGPT (GPT-6 Astra) to assist in developing the proofs of Theorem 1 and the supporting lemmas in November 2026. The author independently verified all mathematical arguments presented in the paper.

Proof of the duality identity

In this appendix, we give a direct proof of Lemma 6 using character sums.

Proof of Lemma 6. Let ω∈C\omega\in\mathbb{C} be a primitive cube root of unity, so that

ω3=1,ω≠1 and 1+ω+ω2=0.\omega^3 = 1, \omega\ne1 \text{ and } 1 + \omega+ \omega^2 = 0.

For an element γ∈F3\gamma\in\mathbb{F}_3, define ωγ=ωϕ(γ)\omega^\gamma= \omega^{\phi(\gamma)}. Since ω3=1\omega^3 = 1, this value is independent of the choice of an integer representative of γ\gamma. For γ,δ∈F3\gamma, \delta\in\mathbb{F}_3, we have

ωγ+δ=ωγωδ,\omega^{\gamma+\delta} = \omega^\gamma\omega^\delta,

since ϕ(γ+δ)≡ϕ(γ)+ϕ(δ)(mod3)\phi(\gamma+\delta) \equiv\phi(\gamma) + \phi(\delta) \pmod{3} and ω3=1\omega^3 = 1.

For a vector x∈F333x \in\mathbb{F}_3^{33}, we have

1∣C∣∑u∈Cωu⋅x={1if x∈C⊥,0if x∉C⊥.(13)\frac{1}{|C|}\sum_{u\in C}\omega^{u\cdot x} = \begin{cases} 1 & \text{if } x \in C^\perp,\\ 0 & \text{if } x \notin C^\perp. \end{cases} \tag*{(13)}

Indeed, if x∈C⊥x \in C^\perp, then ∑u∈Cωu⋅x=∣C∣\sum_{u\in C}\omega^{u\cdot x}=|C|. If x∉C⊥x \notin C^\perp, choose u0∈Cu_0 \in C with u0⋅x≠0u_0 \cdot x \ne0, and let

Qx=∑u∈Cωu⋅x.Q_x = \sum_{u\in C}\omega^{u\cdot x}.

Since the map u↦u+u0u \mapsto u+u_0 is a bijection on CC, we have

Qx=∑u∈Cω(u+u0)⋅x=ωu0⋅xQx.Q_x = \sum_{u\in C}\omega^{(u+u_0)\cdot x} = \omega^{u_0\cdot x}Q_x.

Since ωu0⋅x≠1\omega^{u_0\cdot x} \ne1, it follows that Qx=0Q_x=0. This proves (13).

For x∈{0,1}33x \in\{0,1\}^{33}, let

f(x)=(−1)wt⁡(x)(33−2wt⁡(x)).f(x)=(-1)^{\operatorname{wt}(x)}(33-2\operatorname{wt}(x)).

For a codeword u∈Cu \in C, define

T(u)=∑x∈{0,1}33f(x)ωu⋅x.T(u)=\sum_{x\in\{0,1\}^{33}} f(x)\omega^{u\cdot x}.

From (13) and the definition of S(C⊥)S(C^\perp), we obtain

S(C⊥)=∑x∈C⊥∩{0,1}33f(x)=∑x∈C⊥∩{0,1}33f(x)⋅1+∑x∈{0,1}33∖C⊥f(x)⋅0=∑x∈{0,1}33f(x)(1∣C∣∑u∈Cωu⋅x)=1∣C∣∑u∈C∑x∈{0,1}33f(x)ωu⋅x=1∣C∣∑u∈CT(u).(14)\begin{aligned} S(C^\perp) &= \sum_{x\in C^\perp\cap\{0,1\}^{33}} f(x) \\ &= \sum_{x\in C^\perp\cap\{0,1\}^{33}} f(x)\cdot1+\sum_{x\in\{0,1\}^{33}\setminus C^\perp} f(x)\cdot0 \\ &= \sum_{x\in\{0,1\}^{33}} f(x)\left(\frac{1}{|C|}\sum_{u\in C}\omega^{u\cdot x}\right) \\ &= \frac{1}{|C|}\sum_{u\in C}\sum_{x\in\{0,1\}^{33}} f(x)\omega^{u\cdot x} \\ &= \frac{1}{|C|}\sum_{u\in C}T(u). \tag*{(14)} \end{aligned}

Fix a codeword u=(u1,u2,…,u33)∈Cu=(u_1,u_2,\ldots,u_{33})\in C. For j=0,1,2j=0,1,2, let

mj=∣{i∈{1,2,…,33}∣ui=j}∣.m_j=\left|\{i\in\{1,2,\ldots,33\}\mid u_i=j\}\right|.

Since 1∈C⊆C⊥1\in C\subseteq C^\perp, we have u⋅u=u⋅1=0u\cdot u=u\cdot1=0. Thus, we have

m1+m2≡0(mod3)andm1+2m2≡0(mod3).m_1+m_2\equiv0\pmod{3}\quad\text{and}\quad m_1+2m_2\equiv0\pmod{3}.

Since m0+m1+m2=33m_0+m_1+m_2=33, we obtain

m0≡0(mod3),m1≡0(mod3)andm2≡0(mod3).m_0\equiv0\pmod{3},\quad m_1\equiv0\pmod{3}\quad\text{and}\quad m_2\equiv0\pmod{3}.

Consider the following polynomial:

Hu(z)=∏p=133(1+ωupz)=(1+z)m0(1+ωz)m1(1+ω2z)m2.H_u(z)=\prod_{p=1}^{33}(1+\omega^{u_p}z)=(1+z)^{m_0}(1+\omega z)^{m_1}(1+\omega^2z)^{m_2}.

For each x=(x1,x2,…,x33)∈{0,1}33x=(x_1,x_2,\ldots,x_{33})\in\{0,1\}^{33}, choose 11 from the factor 1+zωup1+z\omega^{u_p} if xp=0x_p=0 and choose zωupz\omega^{u_p} if xp=1x_p=1. The resulting product is

∏1≤p≤33xp=1zωup=zwt⁡(x)ω∑p=133upxp=zwt⁡(x)ωu⋅x.\prod_{\substack{1\leq p\leq33\\x_p=1}}z\omega^{u_p} =z^{\operatorname{wt}(x)}\omega^{\sum_{p=1}^{33}u_px_p} =z^{\operatorname{wt}(x)}\omega^{u\cdot x}.

These choices are in bijection with the vectors x∈{0,1}33x \in\{0,1\}^{33}. Thus, summing over all such choices, we obtain

Hu(z)=∑x∈{0,1}33zwt⁡(x)ωu⋅x.H_u(z) = \sum_{x\in\{0,1\}^{33}} z^{\operatorname{wt}(x)}\omega^{u\cdot x}.

Differentiating Hu(z)H_u(z) and multiplying by zz, we obtain

zHu′(z)=∑x∈{0,1}33wt⁡(x)zwt⁡(x)ωu⋅x.zH_u'(z) = \sum_{x\in\{0,1\}^{33}} \operatorname{wt}(x)z^{\operatorname{wt}(x)}\omega^{u\cdot x}.

Hence, we have

33Hu(z)−2zHu′(z)=∑x∈{0,1}33(33−2wt⁡(x))zwt⁡(x)ωu⋅x.33H_u(z)-2zH_u'(z) = \sum_{x\in\{0,1\}^{33}} \left(33-2\operatorname{wt}(x)\right)z^{\operatorname{wt}(x)}\omega^{u\cdot x}.

Therefore, setting z=−1z=-1 and using the definitions of ff and T(u)T(u), we obtain

T(u)=∑x∈{0,1}33f(x)ωu⋅x=∑x∈{0,1}33(−1)wt⁡(x)(33−2wt⁡(x))ωu⋅x=33Hu(−1)+2Hu′(−1).(15)\begin{aligned} T(u) &= \sum_{x\in\{0,1\}^{33}} f(x)\omega^{u\cdot x} \\ &= \sum_{x\in\{0,1\}^{33}} (-1)^{\operatorname{wt}(x)}\left(33-2\operatorname{wt}(x)\right)\omega^{u\cdot x} \\ &= 33H_u(-1)+2H_u'(-1). \tag*{(15)} \end{aligned}

We consider the two cases m0>0m_0>0 and m0=0m_0=0 separately.

(i) Suppose that m0>0m_0>0. Since m0m_0 is divisible by 3, we have m0≥3m_0\ge3. Thus, Hu(z)H_u(z) is divisible by (1+z)3(1+z)^3, and consequently we have

Hu(−1)=Hu′(−1)=0.H_u(-1)=H_u'(-1)=0.

From (15), we have T(u)=0T(u)=0.

(ii) Suppose that m0=0m_0=0. Let

α=1−ωandβ=1−ω2.\alpha=1-\omega\quad\text{and}\quad\beta=1-\omega^2.

From 1+ω+ω2=01+\omega+\omega^2=0 and ω3=1\omega^3=1, we obtain

β=−ω2α,α2=−3ωandα32=(−3ω)16=316ω.(16)\beta=-\omega^2\alpha,\quad\alpha^2=-3\omega\quad\text{and}\quad\alpha^{32}=(-3\omega)^{16}=3^{16}\omega. \tag*{(16)}

Since m2m_2 is divisible by 3, we have ω2m2=1\omega^{2m_2}=1. Since m0=0m_0=0, we have m1+m2=33m_1+m_2=33. Substituting z=−1z=-1 into Hu(z)H_u(z) and using (16) together with ω2m2=1\omega^{2m_2}=1, we obtain

Hu(−1)=(1−ω)m1(1−ω2)m2=αm1βm2=αm1+m2(α−1β)m2=αm1+m2(−ω2)m2=(−1)m2α33.(17)\begin{aligned} H_u(-1)&=(1-\omega)^{m_1}(1-\omega^2)^{m_2}=\alpha^{m_1}\beta^{m_2}=\alpha^{m_1+m_2}(\alpha^{-1}\beta)^{m_2}\\ &=\alpha^{m_1+m_2}(-\omega^2)^{m_2}=(-1)^{m_2}\alpha^{33}. \tag*{(17)} \end{aligned}

Since m0=0m_0=0, every coordinate of uu is either 1 or 2. Thus, none of the factors of Hu(z)H_u(z) vanishes at z=−1z=-1. Hence, Hu(z)≠0H_u(z)\ne0 in a neighbourhood of −1-1. By the product rule, we have

Hu′(z)=∑p=133ωup∏1≤q≤33q≠p(1+ωuqz).H'_u(z)=\sum_{p=1}^{33}\omega^{u_p}\prod_{\substack{1\le q\le33\\q\ne p}}(1+\omega^{u_q}z).

Dividing both sides by Hu(z)H_u(z) and cancelling common factors, we obtain

Hu′(z)Hu(z)=∑p=133ωup∏1≤q≤33q≠p(1+ωuqz)∏q=133(1+ωuqz)=∑p=133ωup1+ωupz.\frac{H'_u(z)}{H_u(z)} =\sum_{p=1}^{33}\frac{\omega^{u_p}\prod_{\substack{1\le q\le33\\q\ne p}}(1+\omega^{u_q}z)}{\prod_{q=1}^{33}(1+\omega^{u_q}z)} =\sum_{p=1}^{33}\frac{\omega^{u_p}}{1+\omega^{u_p}z}.

Evaluating at z=−1z=-1 and using (16) and (17), we obtain

Hu′(−1)=Hu(−1)(m1ωα+m2ω2β)=(−1)m2α32(m1ω−m2).(18)\begin{aligned} H'_u(-1)&=H_u(-1)\left(\frac{m_1\omega}{\alpha}+\frac{m_2\omega^2}{\beta}\right)\\ &=(-1)^{m_2}\alpha^{32}(m_1\omega-m_2). \tag*{(18)} \end{aligned}

From (15), (16), (17) and (18), we obtain

T(u)=(−1)m2α32(33α+2m1ω−2m2)=(−1)m2316(33ωα+2m1ω2−2m2ω)=(−1)m2316((33−2m2)ω+(2m1−33)ω2)=(−1)m2316(m1−m2)(ω+ω2)=−316(−1)m2(33−2m2).(19)\begin{aligned} T(u)&=(-1)^{m_2}\alpha^{32}(33\alpha+2m_1\omega-2m_2)\\ &=(-1)^{m_2}3^{16}(33\omega\alpha+2m_1\omega^2-2m_2\omega)\\ &=(-1)^{m_2}3^{16}((33-2m_2)\omega+(2m_1-33)\omega^2)\\ &=(-1)^{m_2}3^{16}(m_1-m_2)(\omega+\omega^2)\\ &=-3^{16}(-1)^{m_2}(33-2m_2). \tag*{(19)} \end{aligned}

Here, the fourth equality follows from m1+m2=33m_1+m_2=33, and the last equality follows from ω+ω2=−1\omega+\omega^2=-1 and m1−m2=33−2m2m_1-m_2=33-2m_2.

Since m0=0m_0=0, the vector u+21u+2\mathbf{1} has m1m_1 entries equal to 00 and m2m_2 entries equal to 11. Thus, u+21u+2\mathbf{1} is a (0,1)(0,1)-vector of weight m2m_2. Hence, the definition of ff gives

f(u+21)=(−1)m2(33−2m2).f(u+2\mathbf{1})=(-1)^{m_2}(33-2m_2).

From (19), we obtain

T(u)=−316f(u+21).T(u)=-3^{16}f(u+2\mathbf{1}).

Combining the two cases m0>0m_0>0 and m0=0m_0=0, we have

T(u)={0if wt⁡(u)<33,−316f(u+21)if wt⁡(u)=33.(20)T(u)= \begin{cases} 0 & \text{if }\operatorname{wt}(u)<33,\\ -3^{16}f(u+2\mathbf{1}) & \text{if }\operatorname{wt}(u)=33. \end{cases} \tag*{(20)}

Since 1∈C\mathbf{1}\in C, the map u↦u+21u\mapsto u+2\mathbf{1} sends every codeword of weight 3333 in CC to a vector of B(C)\mathcal{B}(C). Conversely, if x∈B(C)x\in\mathcal{B}(C), then u=x+1u=x+\mathbf{1} belongs to CC and has every coordinate equal to 11 or 22. Thus, we have wt⁡(u)=33\operatorname{wt}(u)=33. These maps are inverse to each other. Hence, u↦u+21u\mapsto u+2\mathbf{1} is a bijection from the set of codewords of weight 3333 in CC onto B(C)\mathcal{B}(C). Using this bijection and (20), we obtain

∑u∈CT(u)=∑u∈Cwt⁡(u)<33T(u)+∑u∈Cwt⁡(u)=33T(u)=0−316∑u∈Cwt⁡(u)=33f(u+21)=−316∑x∈B(C)f(x)=−316S(C).\begin{aligned} \sum_{u\in C}T(u) &=\sum_{\substack{u\in C\\ \operatorname{wt}(u)<33}}T(u)+\sum_{\substack{u\in C\\ \operatorname{wt}(u)=33}}T(u)\\ &=0-3^{16}\sum_{\substack{u\in C\\ \operatorname{wt}(u)=33}}f(u+2\mathbf{1})\\ &=-3^{16}\sum_{x\in\mathcal{B}(C)}f(x)\\ &=-3^{16}S(C). \end{aligned}

From (14), we obtain

S(C⊥)=1∣C∣∑u∈CT(u)=−316∣C∣S(C),S(C^\perp)=\frac{1}{|C|}\sum_{u\in C}T(u)=-\frac{3^{16}}{|C|}S(C),

as required.

References

  1. [1]A.E. Brouwer, Parameters of Strongly Regular Graphs, Available online at http://www.win.tue.nl/~aeb/graphs/srg/srg tab.html.
  2. [2]A.E. Brouwer and H. Van Maldeghem, Strongly Regular Graphs, Encyclopedia Math. Appl., 182, Cambridge University Press, Cambridge, 2022.
  3. [3]D. Ghosh, J. Pal and L.K. Dey, The classification of self-orthogonal codes of length 42, Discrete Math. Algorithms Appl. 10 (2018), 1850083, 8 pp.DOI
  4. [4]M. Harada, A. Munemasa and V.D. Tonchev, Self-dual codes and the nonexistence of a quasi-symmetric 2-(37, 9, 8) design with intersection numbers 1 and 3, J. Combin. Des. 25 (2017), 469–476.DOI
  5. [5]F.J. MacWilliams, C.L. Mallows and N.J.A. Sloane, Generalizations of Gleason’s theorem on weight enumerators of self-dual codes, IEEE Trans. Inform. Theory 18 (1972), 794–805.DOI
  6. [6]A. Munemasa and V.D. Tonchev, Ternary codes, biplanes, and the nonexistence of some quasisymmetric and quasi-3 designs, J. Combin. Des. 28 (2020), 745–752.arxiv.org/abs/2003.04453
  7. [7]A. Munemasa and V.D. Tonchev, Quasi-symmetric 2-(41, 9, 9) designs and doubly even self-dual codes of length 40, Appl. Algebra Engrg. Comm. Comput. 33 (2022), 855–866.DOI
  8. [8]A. Neumaier, “Regular sets and quasi-symmetric 2-designs,” in Combinatorial Theory, D. Jungnickel and K. Vedder (Eds.), Lecture Notes in Math. Vol. 969, 1982, pp. 258–275.DOI
  9. [9]R.M. Pawale, M.S. Shrikhande and K.S. Rajbihar, Non-existence of quasi-symmetric designs with restricted block graphs, Des. Codes Cryptogr. 90 (2022), 871–879.DOI
  10. [10]K.S. Rajbihar and R.M. Pawale, Non-existence of quasi-symmetric designs having certain pseudo-geometric block graphs, AKCE Int. J. Graphs Comb. 21 (2024), 120–128.
  11. [11]S. Shpectorov and T. Zhao, Strongly regular graphs with parameters (85, 14, 3, 2) do not exist, J. Combin. Theory Ser. A 225 (2027), 106242.arxiv.org/abs/2504.02449
  12. [12]M.S. Shrikhande, “Quasi-symmetric designs,” in Handbook of Combinatorial Designs (second edition), C.J. Colbourn and J.H. Dinitz (Eds.), Chapman & Hall/CRC, Boca Raton, FL, 2007, pp. 578–582.
  13. [13]M.S. Shrikhande and S.B. Sane, Quasi-Symmetric Designs, London Math. Soc. Lecture Note Ser. 164, Cambridge University Press, Cambridge, 1991.DOI
  14. [14]V.D. Tonchev, “Codes,” in Handbook of Combinatorial Designs (second edition), C.J. Colbourn, J.H. Dinitz (Eds.), Chapman & Hall/CRC, Boca Raton, FL, 2007, pp. 677–702.

Paper details

vlnc:W_8k5p6awp
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

ChatGPT

View statement
Looking for related papers…
Contents