Hardness of Quantum Distribution Learning and Quantum Cryptography

TL;DR

This paper establishes a fundamental link between the existence of quantum one-way puzzles and the hardness of quantum distribution learning, revealing deep connections between quantum cryptography and computational learning theory.

Contribution

It provides the first complete characterization of quantum one-way puzzles in terms of the hardness of proper quantum distribution learning.

Findings

OWPs exist iff proper quantum distribution learning is hard on average.

Worst-case hardness of quantum distribution learning from PP ≠ BQP is unlikely without complexity class collapses.

PP ≠ BQP is equivalent to the hardness of agnostic quantum distribution learning.

Abstract

The existence of one-way functions (OWFs) forms the minimal assumption in classical cryptography. However, this is not necessarily the case in quantum cryptography. One-way puzzles (OWPuzzs), introduced by Khurana and Tomer, provide a natural quantum analogue of OWFs. The existence of OWPuzzs implies $PP\neq BQP$, while the converse remains open. In classical cryptography, the analogous problem-whether OWFs can be constructed from $P \neq NP$-has long been studied from the viewpoint of hardness of learning. Hardness of learning in various frameworks (including PAC learning) has been connected to OWFs or to $P \neq NP$. In contrast, no such characterization previously existed for OWPuzzs. In this paper, we establish the first complete characterization of OWPuzzs based on the hardness of a well-studied learning model: distribution learning. Specifically, we prove that OWPuzzs exist if and only if proper quantum distribution learning is hard on average. A natural question that follows is whether the worst-case hardness of proper quantum distribution learning can be derived from $PP \neq BQP$. If so, and a worst-case to average-case hardness reduction is achieved, it would imply OWPuzzs solely from $PP \neq BQP$. However, we show that this would be extremely difficult: if worst-case hardness is PP-hard (in a black-box reduction), then $SampBQP \neq SampBPP$ follows from the infiniteness of the polynomial hierarchy. Despite that, we show that $PP \neq BQP$ is equivalent to another standard notion of hardness of learning: agnostic. We prove that $PP \neq BQP$ if and only if agnostic quantum distribution learning with respect to KL divergence is hard. As a byproduct, we show that hardness of agnostic quantum distribution learning with respect to statistical distance against $PPT^{\Sigma_3^P}$ learners implies $SampBQP \neq SampBPP$.