基于MZ调制器伊辛机的网络最大割优化算法

打开文本图片集

中图分类号：0437.4 文献标志码：A doi：10.12415/j.issn.1671-7872.24061

Network Max-cut Optimization Algorithm Based on MZ Modulator Ising Machine

LU Pingping， CHEN Zhile， DING Chengxiang， WANG Dong （School of Microelectronics & Data Science，Anhui University of Technology，Maanshan ， China）

Abstract：To address the isse of exponential growth in computation time with problem size in traditional enumeration algorithms for solving combinatorial optimization problems， a simulated Ising machine algorithm based on Mach-Zehnder （MZ） modulators was proposed.The interference output characteristics of the MZ modulator were modeled using a cosine function，and Gaussan random noise was introduced to simulate system noise，combined with numerical iteration to simulate the system's dynamic evolution. Experiments were conducted on max-cut problems in regular，small-world，and random networks with16 and10o vertices.The results demonstrate that a success rate of 100% is achieved for 16-vertex networks，while an 88% success rate is maintained for 1oo-vertex random networks. In terms of computational eficiency， the Ising algorithm requires only 0.42 s for solving a 25- vertex regular network，showing a significant advantage over the enumeration method （29.93 s）.This study provides aneffcient solution forcomplex network optimization problemsand oferstheoretical references for the experimental design of MZ modulator-based Ising machines.

Keywords：combination optimization; graph; Ising machine; maximum cut; smal-world network; regular network; random network

伊辛模型作为描述自旋相变的经典数学模型，通过将物理系统的振幅、相位和偏振等特性映射到自旋模型，为构建求解组合优化问题的伊辛机提供了理论基础。（剩余13298字）