Parallel Quantum Local Search via Evolutionary Mechanism

TL;DR

This paper introduces a Parallel Quantum Local Search method that uses small quantum computers to solve complex optimization problems more efficiently by running multiple search pathways simultaneously and combining their best results.

Contribution

It presents a novel parallel approach to quantum local search that overcomes sequential limitations, accelerating convergence in solving Ising and combinatorial optimization problems.

Findings

Enhanced solution quality for Ising problems

Significant acceleration in convergence speed

Effective utilization of small-scale quantum computers

Abstract

We propose an innovative Parallel Quantum Local Search (PQLS) methodology that leverages the capabilities of small-scale quantum computers to efficiently address complex combinatorial optimization problems. Traditional Quantum Local Search (QLS) methods face limitations due to the sequential nature of solving sub-problems, which arises from dependencies between their solutions. Our approach transcends this constraint by simultaneously executing multiple QLS pathways and aggregating their most effective outcomes at certain intervals to establish a ``generation''. Each subsequent generation commences with the optimal solution from its predecessor, thereby significantly accelerating the convergence towards an optimal solution. Our findings demonstrate the profound impact of parallel quantum computing in enhancing the resolution of Ising problems, which are synonymous with combinatorial optimization challenges.