Solving constrained slot placement problems using an ising machine and its evaluations

Sho Kanamaru, Kazushi Kawamura, Shu Tanaka, Yoshinori Tomita, Nozomu Togawa

Research output: Contribution to journalArticlepeer-review

Abstract

Ising machines have attracted attention, which is expected to obtain better solutions of various combinatorial optimization problems at high speed by mapping the problems to natural phenomena. A slot-placement problem is one of the combinatorial optimization problems, regarded as a quadratic assignment problem, which relates to the optimal logic-block placement in a digital circuit as well as optimal delivery planning. Here, we propose a mapping to the Ising model for solving a slot-placement problem with additional constraints, called a constrained slot-placement problem, where several item pairs must be placed within a given distance. Since the behavior of Ising machines is stochastic and we map the problem to the Ising model which uses the penalty method, the obtained solution does not always satisfy the slot-placement constraint, which is different from the conventional methods such as the conventional simulated annealing. To resolve the problem, we propose an interpretation method in which a feasible solution is generated by post-processing procedures. We measured the execution time of an Ising machine and compared the execution time of the simulated annealing in which solutions with almost the same accuracy are obtained. As a result, we found that the Ising machine is faster than the simulated annealing that we implemented.

Original languageEnglish
Pages (from-to)226-236
Number of pages11
JournalIEICE Transactions on Information and Systems
VolumeE104D
Issue number2
DOIs
Publication statusPublished - 2021 Feb 1

Keywords

  • Ising machine
  • Ising model
  • Quadratic assignment problem
  • Quadratic unconstrained binary optimization
  • Slot-placement problem

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Artificial Intelligence

Fingerprint Dive into the research topics of 'Solving constrained slot placement problems using an ising machine and its evaluations'. Together they form a unique fingerprint.

Cite this