Scholarlink

<p>Fifth-generation (5G) wireless cellular networks have attracted considerable attention due to emerging services and their ability to meet diverse infrastructure requirements. Network slicing has been introduced in 5G to address heterogeneous user service requirements. Each slice operates independently and delivers tailored services to users. Various techniques have been proposed to automate network slicing functions and optimize quality of service. However, these approaches incur significant computational overhead when the number of actions and system states increases. This study offers a unified deep learning design, the Xception Convolutional Belief Forward Harmonic Network (XConBFHNet), for 5G network slicing. The system model considers a sliced 5G network, where the physical infrastructure is accessed through the Radio Access Network (RAN) and features are collected from multiple devices. These features are weighted using Spider Wasp Optimizer (SWO), and the network slices are subsequently classified using XConBFHNet. The proposed model achieved an average response time of 0.274&#xa0;s, latency of 0.269&#xa0;s, availability of 92.71%, and energy efficiency of 93.025 Mbps/J.</p>

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 hybrid Xception-belief forward harmonic network for resource allocation in 5G network slicing

  • 期刊论文

  • 发表日期：

    2026年1月5日

查看全文
  • Kigninman Désiré Kone,
  • Adlès Francis Kouassi,
  • Tanon Lambert Kadjo,
  • Nabil Tabbane,
  • Olivier Asseu

摘要

Fifth-generation (5G) wireless cellular networks have attracted considerable attention due to emerging services and their ability to meet diverse infrastructure requirements. Network slicing has been introduced in 5G to address heterogeneous user service requirements. Each slice operates independently and delivers tailored services to users. Various techniques have been proposed to automate network slicing functions and optimize quality of service. However, these approaches incur significant computational overhead when the number of actions and system states increases. This study offers a unified deep learning design, the Xception Convolutional Belief Forward Harmonic Network (XConBFHNet), for 5G network slicing. The system model considers a sliced 5G network, where the physical infrastructure is accessed through the Radio Access Network (RAN) and features are collected from multiple devices. These features are weighted using Spider Wasp Optimizer (SWO), and the network slices are subsequently classified using XConBFHNet. The proposed model achieved an average response time of 0.274 s, latency of 0.269 s, availability of 92.71%, and energy efficiency of 93.025 Mbps/J.