Table of Contents
Fetching ...

On the minimal length of addition chains

Jean-Marie De Koninck, Nicolas Doyon, William Verreault

TL;DR

This paper studies the distribution of the minimal addition-chain length by introducing $F(m,r)$, the count of integers in $[2^m,2^{m+1})$ with $\ell(n)\le m+r$. It provides precise upper and lower bounds for $F(m,\tfrac{cm}{\log m})$ in terms of $cm$ and subpolynomial corrections, extending Erdős’s result on $\ell(n)$ for almost all $n$. The authors develop a detailed combinatorial framework that partitions addition steps into doubling, large, midsize, and small types, analyzes their interdependencies via block structures, and leverages marked-block arguments to bound the number of candidate chains. A constructive lower-bound argument using binary expansion yields matching exponential-scale growth up to logarithmic corrections. Together, these results push toward a fuller understanding of how the minimal-length addition chains are distributed across integers, while highlighting remaining gaps between upper and lower estimates.

Abstract

We denote by $\ell(n)$ the minimal length of an addition chain leading to $n$ and we define the counting function $$ F(m,r):=\#\left\{n\in[2^m, 2^{m+1}):\ell(n)\le m+r\right\}, $$ where $m$ is a positive integer and $r\ge 0$ is a real number. We show that for $0< c<\log 2$ and for any $\varepsilon>0$, we have as $m\to \infty$, $$ F\left(m,\frac{cm}{\log m}\right)<\exp\left(cm+\frac{\varepsilon m\log\log m}{\log m}\right) $$ and $$ F\left(m,\frac{cm}{\log m}\right)>\exp\left(cm-\frac{(1+\varepsilon)cm\log\log m}{\log m}\right). $$ This extends a result of Erdős which says that for almost all $n$, as $n\to\infty$, $$ \ell(n)=\frac{\log n}{\log 2}+\left(1+o(1)\right)\frac{\log n}{\log \log n}. $$

On the minimal length of addition chains

TL;DR

This paper studies the distribution of the minimal addition-chain length by introducing , the count of integers in with . It provides precise upper and lower bounds for in terms of and subpolynomial corrections, extending Erdős’s result on for almost all . The authors develop a detailed combinatorial framework that partitions addition steps into doubling, large, midsize, and small types, analyzes their interdependencies via block structures, and leverages marked-block arguments to bound the number of candidate chains. A constructive lower-bound argument using binary expansion yields matching exponential-scale growth up to logarithmic corrections. Together, these results push toward a fuller understanding of how the minimal-length addition chains are distributed across integers, while highlighting remaining gaps between upper and lower estimates.

Abstract

We denote by the minimal length of an addition chain leading to and we define the counting function where is a positive integer and is a real number. We show that for and for any , we have as , and This extends a result of Erdős which says that for almost all , as ,

Paper Structure

This paper contains 14 sections, 16 theorems, 136 equations.

Key Result

Theorem 3.1

For any $\varepsilon>0$ and for $0< c<\log 2$, we have for $m$ large enough

Theorems & Definitions (41)

  • Definition 1.1
  • Definition 1.2
  • Definition 1.3
  • Theorem 3.1
  • Theorem 3.2
  • Definition 4.1
  • Definition 4.2
  • Lemma 4.1
  • proof
  • Lemma 4.2
  • ...and 31 more