Two-sources Randomness Extractors for Elliptic Curves

TL;DR

This paper introduces new functions for elliptic curves that extract nearly uniform random bits from two independent points, enhancing cryptographic applications like key exchange and pseudo-random number generation.

Contribution

It extends existing elliptic curve randomness extractors to two-source scenarios, providing novel constructions for extracting bits from combined points over finite fields.

Findings

Extracted bits are close to uniform distribution.

Functions work for both prime and binary fields.

Extends previous single-source extractors to two-source setting.

Abstract

This paper studies the task of two-sources randomness extractors for elliptic curves defined over finite fields $K$, where $K$ can be a prime or a binary field. In fact, we introduce new constructions of functions over elliptic curves which take in input two random points from two differents subgroups. In other words, for a ginven elliptic curve $E$ defined over a finite field $\mathbb{F}_q$ and two random points $P \in \mathcal{P}$ and $Q\in \mathcal{Q}$, where $\mathcal{P}$ and $\mathcal{Q}$ are two subgroups of $E(\mathbb{F}_q)$, our function extracts the least significant bits of the abscissa of the point $P\oplus Q$ when $q$ is a large prime, and the $k$-first $\mathbb{F}_p$ coefficients of the asbcissa of the point $P\oplus Q$ when $q = p^n$, where $p$ is a prime greater than $5$. We show that the extracted bits are close to uniform. Our construction extends some interesting randomness extractors for elliptic curves, namely those defined in \cite{op} and \cite{ciss1,ciss2}, when $\mathcal{P} = \mathcal{Q}$. The proposed constructions can be used in any cryptographic schemes which require extraction of random bits from two sources over elliptic curves, namely in key exchange protole, design of strong pseudo-random number generators, etc.