Infinite families of \(t\)-designs and strongly regular graphs from punctured codes (Q6112185)

Let \(\mathcal{P}\) be a set with \(v \geq 1\) elements and \(\mathcal{B}\) a set of \(k\)-subsets of \(\mathcal{P}\) with \(1 \leq k \leq v\). The elements in \(\mathcal{P}\) are called points and the elements in \(\mathcal{B}\) are said to be blocks. Let \(t\) be a positive integer where \(t \leq k\). If every \(t\)-subset in \(\mathcal{P}\) is contained in exactly \(\lambda\) elements of \(\mathcal{B}\), we call the pair \(\mathcal{D} = (\mathcal{P}, \mathcal{B})\) a \(t\)-\((v,k,\lambda)\) design, or simply \(t\)-design. A connected graph \(\Gamma=(V,E)\) is called a strongly regular graph with parameters \((n,k,\lambda,\mu)\) if \(\Gamma\) is regular with valency \(k\) and satisfying the properties that for any given two different vertices, they have \(\lambda\) or \(\mu\) common neighbors according as the two given vertices are adjacent or not. An \([n,k,d]\) linear code over (a finite field) \(\mathbb{F}_q\) is a \(k\)-dimensional linear subspace of \(\mathbb{F}_q^n\) with minimum Hamming distance \(d(C)=d\). The dual of an \([n, k]\) linear code \(C\) over \(\mathbb{F}_q\) is defined as \(C^\perp:=\{\mathbf{d} \in \mathbb{F}_q^n:~ \mathbf{d} \cdot \mathbf{c}=0, \forall \mathbf{c} \in C\}\). The code \(C\) is said to be projective if the minimum Hamming distance of its dual, \(d(C^\perp)\), is at least \(3\). Constructing codes from the old ones is one important issue in coding theory. Let \(C\) be a linear code of parameters \([n,k,d]\) and \(T \subseteq \{1,2,\ldots,n\}\). A punctured code \(C^T\) is a code obtained from \(C\) by deleting all coordinates in \(T\). It means that the punctured code \(C^T\) is of length \(n-|T|\). The work in the paper under review is motivated by two facts: \begin{itemize} \item[(1)] By the celebrated Assmus-Mattson theorem [\textit{E. F. Assmus jun.} and \textit{H. F. Mattson jun.}, SIAM Rev. 16, 349--388 (1974; Zbl 0268.94002)], we have to construct a linear code \(C\) with only a few Hamming weights whose dual \(C^\perp\) has large minimum distance. In other words, projective linear codes with only a few weights are very promising in holding \(t\)-designs. \item[(2)] \textit{R. Calderbank} and \textit{W. M. Kantor} [Bull. Lond. Math. Soc. 18, 97--122 (1986; Zbl 0582.94019)] established the relationship between projective two-weight codes and certain strongly regular graphs. They proved that projective two-weight codes over \(\mathbb{F}_q\) yield strongly regular graphs. \end{itemize} Since projective codes with a few weights are useful in constructing \(t\)-deigns and strongly regular graphs, the paper tries to construct projective linear codes with only a few nonzero Hamming weights. Among the main results of the paper are as follows: (1) To construct the family of three-weight projective codes punctured from reducible cyclic codes (Lemma 3.1, Theorem 3.2, and Corollary 1.) and obtain infinite families of \(2\)-designs (Theorem 3.3). (2) To construct the family of two-weight projective codes punctured from a class of linear codes (Lemma 4.1, Lemma 4.2, and Theorem 4.3). Then infinite families of \(3\)-designs including Steiner systems are derived (Theorem 4.4). (3) To construct the family of one-weight and two-weight projective codes punctured from irreducible cyclic codes (Theorem 5.2 and Theorem 5.3) and derive infinite families of \(2\)-designs and \(3\)-designs including Steiner systems (Theorem 5.5 and Theorem 5.6). (4) To construct a new family of projective two-weight codes over \(\mathbb{F}_3\) (Theorem 6.4) and (5) To use all projective two-weight codes in the paper to construct strongly regular graphs (Section 6.2 of the paper).