Detuning Choice for solving MIS and MWIS

TL;DR

This paper introduces a new detuning computation method to mitigate parasitic interactions in quantum MWIS and MIS problems, tailored for current Pasqal neutral-atom quantum hardware, with three variants tested on emulators.

Contribution

A novel detuning computation approach for quantum MWIS and MIS problems that adapts to different hardware maturity levels and mitigates parasitic interactions.

Findings

All three variants perform effectively on graphs up to 30 qubits.

The methods are compatible with current hardware constraints.

Results demonstrate realistic and transferable performance.

Abstract

We address the realization of Maximum Weighted Independent Set (MWIS) and quantum Maximum Independent Set (MIS) problems under Pasqal's neutral-atom QPU constraints: limited qubit number, bounds on {\Omega} and {\Delta}, sequence duration, confinement space, minimum distances, and parasitic interactions. To obtain results directly compatible with current hardware, we propose a new detuning computation method that mitigates parasitic interactions in arbitrary asymmetric graphs. Three variants are introduced, matching different hardware maturity levels: (I) a theoretical local-detuning approach; (II) a Detuning Map Modulation (DMM) method bridging theory and near-term feasibility; and (III) a global-pulse implementation suitable for current QPUs. All were evaluated on Pasqal's emulators for graphs up to 30 qubits, demonstrating realistic, transferable performance within present-day hardware limits.