Generalized Additive Bases in Number Theory

• Generalized additive bases are subsets of numbers or group elements where every element is representable as a sum of a fixed number of members, extending the classical additive basis concept.
• Methodologies using the Stein–Chen method and Janson's inequalities yield precise probabilistic thresholds and asymptotic results for complete additive coverings.
• Applications range from abelian group constructions to combinatorial designs in coding theory and signal processing, highlighting the versatile impact of these bases.

A generalized additive basis is a concept from additive combinatorics and number theory that extends the classical notion of an additive basis by considering representations of elements in a finite set or group as sums of a fixed number of elements from a subset, possibly with additional constraints. Traditionally, a set $A \subset \{0,1,\ldots,n\}$ is a 2-additive basis for $\{1,\ldots,n\}$ if every $j$ in $\{1,\ldots,n\}$ can be expressed as $j = x + y$ with $x, y \in A$. The notion broadens to $k$-additive bases, where every element is expressible as the sum of $k$ members of $A$. Beyond this, modular and truncated variants address representation modulo $n$ or only over central intervals to mitigate boundary effects. Theoretical developments connect these bases to probabilistic models, combinatorial design, and algebraic structures such as abelian groups and finite fields.

1. Classical and Generalized Definitions

The classical definition considers $A \subset [n] \cup \{0\}$ as a 2-additive basis for $[n]$ if $\forall j \in [n]$, $j = x + y$ for some $x, y \in A$; for $k$-additive bases, the expression generalizes to $j = x_1 + \cdots + x_k$, $x_i \in A$ ($i=1,\ldots,k$). Modular additive bases refer to representation modulo $n$, and truncated bases restrict representation to central intervals $[an, (k-a)n]$ with $0 < a < 1$, thereby diminishing forced boundary terms.

In group-theoretic settings, a $k$-additive basis of a finite abelian group $G$ is a multiset $A$ such that every $g \in G$ can be written as $g = a_1 + \cdots + a_k$ with each $a_i \in A$. For a $g$-additive basis (as in (Li et al., 28 Sep 2025)), every element of $G$ must be representable as a sum of two elements from $A$ in at least $g$ distinct ways; $g$-difference bases analogously address differences.

2. Probabilistic Thresholds and Asymptotics

Random selection of basis elements leads to sharp threshold phenomena for the emergence of additive bases. For $A$ constructed by independent inclusion of each element of $[n] \cup \{0\}$ with $p=p_n \to 0$, the probability of forming a 2-additive basis exhibits a sudden change at the threshold $p = \sqrt{(1/2 \log n - 2 \log\log n + A_n)/n}$ (Godbole et al., 2011). At the transition, the probability approaches $\exp(-2a e^{-A/2})$ for truncated bases. For modular bases, the threshold is $p = (1/(2 \log n) + A_n/n)$. For $k$-additive bases, the asymptotic threshold is $$p = \left( (k!\log n + o(\log n)) / n^{k-1} \right)^{1/k}$$.

These results are established using the Stein–Chen method for Poisson approximation—showing that the count of uncovered indices $X$ is asymptotically Poisson distributed—and Janson's inequalities to quantify dependencies among representation events, thus offering precise control over the emergence of full additive bases in random models.

3. Additive Bases in Abelian Groups and Semigroups

The study extends to infinite and finite abelian groups, introducing analogues of classical functions:

• $E_G(h)$: Maximum number of exceptional elements in a basis of order $h$; upper bounded by $h - 1$ for any infinite $G$ (Lambert et al., 2015).
• $X_G(h)$: Maximal order of a basis after removal of a regular element; quadratic growth for groups with a cyclic quotient (e.g., $X_G(h) \geq (h(h+4))/3$) but linear for groups whose elements have bounded order (e.g., $X_G(h) \leq ph + p - 1$ for torsion group of order $p$).
• $S_G(h)$: Minimum $s$ such that, for any basis $A$ of order $h$, removal of almost any $a \in A$ leaves a basis of order $\le s$; $h+1 \leq S_G(h) \leq 2h$.

Analyses employ combinatorial construction (e.g., "digit expansions"), amenability/invariant means, and algebraic group properties. The finiteness of essential subsets in bases and sharp bounds on $X_G(h,k)$ ($O(h^{2k+1})$ for $k$ elements removed) (Bienvenu et al., 2020) generalizes results from integers to translatable semigroups.

4. Explicit Bounds in Finite Group Products

In the context of $g$-additive and $g$-difference bases for cartesian products $G^n$, explicit upper bounds for their minimal sizes arise from combinatorial and algebraic constructions (Li et al., 28 Sep 2025). For groups $Z_{p^s}^{2n}$ (excluding small cases), one obtains:

• For $g=1,2$: $$\nu_1(Z_{p^s}^{2n}) \leq \nu_2(Z_{p^s}^{2n}) \leq 2(p^{sn} - 1)$$.
• For $g=3,4,5,6$: $$\nu_3(Z_{p^s}^{2n}) \leq \nu_4(Z_{p^s}^{2n}) \leq \nu_5(Z_{p^s}^{2n}) \leq \nu_6(Z_{p^s}^{2n}) \leq 3(p^{sn} - 1)$$.
• The corresponding difference basis bounds $\eta_g$ match these up to small additive corrections.

The underlying constructions utilize combinatorial designs, Galois ring machinery, and direct product decompositions, and they can be iterated for larger $n$, e.g., $$\nu_1(Z_{p^s}^{2n}) \leq 4(p^s - 1)(p^{s(n-1)} - 1)$$ for $n \geq 2$.

g	Minimal size $\nu_g(Z_{p^s}^{2n})$	Minimal size $\eta_g(Z_{p^s}^{2n})$
1-2	$\leq 2(p^{sn} - 1)$	$\leq 2(p^{sn} - 1)$
3-6	$\leq 3(p^{sn} - 1)$	$\leq 3(p^{sn} - 1)$

This suggests an approximate proportionality between $g$ and the minimal basis size, with the coefficient determined by specific combinatorial coverings.

5. Connections to Design Theory and Applications

The construction of $g$-additive bases in $G^n$ draws on combinatorial design ideas—relative difference sets, Galois rings, and Teichmüller systems furnish building blocks matching sum and difference constraints. These bases find relevance both in pure mathematics and in applied areas such as array signal processing (cf. planar additive bases (Kohonen et al., 2017)) and coding theory, where covering properties translate into performance guarantees.

In probabilistic and algorithmic settings, these structural results provide input for random constructions, efficient enumeration, and lower/upper bound analyses for combinatorial structures with prescribed additive covering properties.

6. Broader Framework and Research Directions

Generalized additive bases underpin quantitative results on representation functions, threshold phenomena, and robustness under perturbation. They connect to open questions including efficient coset covers, additive bases in groups and semigroups, extremal combinatorial optimization, and deeper algebraic invariants. Methodologically, Poisson approximation, Janson's inequalities, invariant mean techniques, and combinatorial design theory supply critical tools. Ongoing work aims to sharpen bounds, link finite and asymptotic regimes, and apply these bases in both analytic and combinatorial contexts.

A plausible implication is that continuing refinement of construction techniques—via richer algebraic structure, combinatorial optimization, and probabilistic analysis—will yield even sharper bounds and new applications for generalized additive bases in a range of mathematical disciplines.