An Approximate Parallel Annealing Ising Machine for Solving Traveling Salesman Problems

Annealing-based Ising machines have emerged as high-performance solvers for combinatorial optimization problems (COPs). As a typical COP with constraints imposed on the solution, traveling salesman problems (TSPs) are difficult to solve using conventional methods. To address this challenge, we desig...

Full description

Saved in:
Bibliographic Details
Published in:IEEE embedded systems letters 2023-12, Vol.15 (4), p.226-229
Main Authors: Tao, Qichao, Zhang, Tingting, Han, Jie
Format: Article
Language:English
Subjects:
Algorithms
Annealing
Combinatorial analysis
Constraints
Floating point arithmetic
Hardware
Heuristic algorithms
Indexes
Integrated circuit modeling
Ising model
parallel annealing (PA)
Registers
Scaling factors
Simulated annealing
simulated annealing (SA)
Traveling salesman problem
traveling salesman problem (TSP)
Traveling salesman problems
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Annealing-based Ising machines have emerged as high-performance solvers for combinatorial optimization problems (COPs). As a typical COP with constraints imposed on the solution, traveling salesman problems (TSPs) are difficult to solve using conventional methods. To address this challenge, we design an approximate parallel annealing Ising machine (APAIM) based on an improved parallel annealing algorithm. In this design, adders are reused in the local field accumulator units (LAUs) with half-precision floating-point representation of the coefficients in the Ising model. The momentum scaling factor is approximated by a linear, incremental function to save hardware. To improve the solution quality, a buffer-based energy calculation unit selects the best solution among the found candidate results in multiple iterations. Finally, approximate adders are applied in the design for improving the speed of accumulation in the LAUs. The design and synthesis of a 64-spin APAIM show the potential of this methodology in efficiently solving complicated constrained COPs.
ISSN:1943-0663
1943-0671
DOI:10.1109/LES.2023.3298739