Scholarlink

Optimizing router placement in Wireless Mesh Networks (WMNs) is essential for reliable service delivery, yet it constitutes a computationally challenging because is a NP-hard problem. Extending previous small-scale simulations, this research introduces a hybrid system for medium-scale WMNs to address this challenge, specifically simulating a mesh network containing 96 clients in two-island mesh client distribution. We utilized WMN-PSOHCDGA system, which integrates Particle Swarm Optimization (PSO), Hill Climbing (HC), and a Distributed Genetic Algorithm (DGA). For performance evaluation, we compared two PSO router replacement methods: Random Inertia Weight Method (RIWM) and Fast Convergence Rational Decreasing Vmax Method (FC-RDVM), and two crossover methods: Simplex Crossover (SPX) and Unimodal Normal Distribution Crossover (UNDX). All evaluated scenarios resulted in full mesh router interconnection. However, the combination of SPX and FC-RDVM provided the best network client connectivity (NCMC), while the pairing of UNDX and RIWM provided the most effective load balancing.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Comparison Study of SPX and UNDX Crossover Methods Considering RIWM and FC-RDVM Router Replacement Methods for Middle-Scale WMN and Two Islands Distribution of Mesh Clients

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Paboth Kraikritayakul,
  • Admir Barolli,
  • Yusuke Irie,
  • Phudit Ampririt,
  • Shinji Sakamoto,
  • Leonard Barolli

摘要

Optimizing router placement in Wireless Mesh Networks (WMNs) is essential for reliable service delivery, yet it constitutes a computationally challenging because is a NP-hard problem. Extending previous small-scale simulations, this research introduces a hybrid system for medium-scale WMNs to address this challenge, specifically simulating a mesh network containing 96 clients in two-island mesh client distribution. We utilized WMN-PSOHCDGA system, which integrates Particle Swarm Optimization (PSO), Hill Climbing (HC), and a Distributed Genetic Algorithm (DGA). For performance evaluation, we compared two PSO router replacement methods: Random Inertia Weight Method (RIWM) and Fast Convergence Rational Decreasing Vmax Method (FC-RDVM), and two crossover methods: Simplex Crossover (SPX) and Unimodal Normal Distribution Crossover (UNDX). All evaluated scenarios resulted in full mesh router interconnection. However, the combination of SPX and FC-RDVM provided the best network client connectivity (NCMC), while the pairing of UNDX and RIWM provided the most effective load balancing.