Scholarlink

Wireless Mesh Networks (WMNs) offer a robust and adaptable network across a large region by connecting mesh routers. Since mesh node placement directly affects network coverage, connectivity, and performance, it is one of the most challenging problems in WMNs. This is classified as NP-Hard problem. To address this problem, we consider the application of intelligent algorithms such as GA (Genetic Algorithms) and PSO (Particle Swarm Optimization). We combine these algorithms and implement WMN-PSODGA simulation system. We carry out simulations for different router replacement methods (CM, LDVM, RDVM) and two crossover Methods (UNDX and UNDX-m) considering three metrics: SGC, NCMC, and NCMCpR. The simulation results show that the network connectivity is good because SGC was 100% for CM, LDVM, and RDVM router replacement methods and UNDX and UNDX-m. However, the client coverage was not good because NCMC was about 50%. Considering load balancing, the combination of UNDX with LDVM has the best performance compared with other combinations of router replacement methods and crossover methods.

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 Three Router Replacement Methods for a Middle-Scale WMN Considering UNDX and UNDX-m Crossover Methods and Boulevard Distribution of Mesh Clients

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yusuke Irie,
  • Paboth Kraikritayakul,
  • Shinji Sakamoto,
  • Makoto Ikeda,
  • Keita Matsuo,
  • Leonard Barolli

摘要

Wireless Mesh Networks (WMNs) offer a robust and adaptable network across a large region by connecting mesh routers. Since mesh node placement directly affects network coverage, connectivity, and performance, it is one of the most challenging problems in WMNs. This is classified as NP-Hard problem. To address this problem, we consider the application of intelligent algorithms such as GA (Genetic Algorithms) and PSO (Particle Swarm Optimization). We combine these algorithms and implement WMN-PSODGA simulation system. We carry out simulations for different router replacement methods (CM, LDVM, RDVM) and two crossover Methods (UNDX and UNDX-m) considering three metrics: SGC, NCMC, and NCMCpR. The simulation results show that the network connectivity is good because SGC was 100% for CM, LDVM, and RDVM router replacement methods and UNDX and UNDX-m. However, the client coverage was not good because NCMC was about 50%. Considering load balancing, the combination of UNDX with LDVM has the best performance compared with other combinations of router replacement methods and crossover methods.