A note on weighted consecutive Davenport constant

TL;DR

The paper studies the weighted consecutive Davenport constant \(C_A(G)\), proving \(C_A(C_n^2)=4\) for full weight \(A=[1,n-1]\), \(C(G)=|G|\) for finite abelian and metacyclic groups, and \(C(H\times K)=|H||K|\) when \(H\) is finite abelian and \(K\) is metacyclic. It also establishes product lower bounds and extends weighted unit-weight results to higher-rank groups.

Key findings

• For finite abelian \(G\), \(C(G)=|G|\) (Theorem 3.1); for metacyclic \(G\), \(C(G)=|G|\) (Theorem 3.2); consequently, if \(G=H\times K\) with \(H\) finite abelian and \(K\) metacyclic, then \(C(G)=|H||K|\) (Corollary 3.3).
• For full weight \(A=[1,n-1]\), \(C_A(C_n^2)=4\) (stated as \(C_A(C_2n)=4\) in the abstract; Theorem 4.4), and more generally \(C_A(C_n^r)\ge 2^r\) (Proposition 4.2(a)).
• Product lower bound: if \(H,K\) are finite groups with \(\exp(H)=\exp(K)\) and \(A\subseteq[1,\exp(K)-1]\), then \(C_A(H\times K)\ge C_A(H)C_A(K)\); for \(A=\{1\}\), the \(\exp(H)=\exp(K)\) hypothesis can be removed (Proposition 2.3).
• For odd prime \(p\), \(A=U(p)^2\) or \(A=U(p)\setminus U(p)^2\) gives \(C_A(C_p)=3\) and \(C_A(C_p^r)\ge 3^r\); for \(p\equiv1\pmod3\), \(p\ne7\), \(A=U(p)\) gives \(C_A(C_p)=3\) (Lemma 4.1, Proposition 4.2). Lemma 4.5 gives \(D_A(G)=r+1\) under stated conditions.

Methods

• Theoretical combinatorial number theory: defines ordered sequences over a multiplicative group, \(A\)-weighted consecutive products, product-one free sequences, and the constant \(C_A(G)\).
• Proofs use induction on rank, explicit lower-bound constructions of product-one free sequences, and upper bounds via Lemma 2.1 (any sequence of length \(\ge |G|\) has a consecutive product-one subsequence) plus prior weighted Davenport results.
• No empirical data or sample; the work consists of definitions, lemmas, propositions, and theorems with proofs.

Limitations

• The paper has no dedicated limitations section; it notes that the ordinary Davenport constant \(D(G)\) is still unknown for a wide class of groups and poses an open problem.
• Evident from the text: several displayed formulas are omitted/undecoded in the provided source, so some theorem statements (e.g., Theorem 4.6) cannot be fully read here.
• Evident from the text: several results carry restrictive hypotheses (e.g., \(\exp(H)=\exp(K)\), \(\gcd(d,n)\le d-1\), \(d^k n\ge6\)), limiting their scope.