6G+ Networks Through Enhanced Efficiency and Sustainability With MADDPG-Driven Network Slicing in SoS Environments

Research output: Contribution to journalArticlepeer-review

Abstract

This study explores the integration of sustainable practices in the advancing domain of sixth-generation and beyond (6G+) network technologies, with a particular focus on enhancing the efficiency of search and rescue operations. It presents a comprehensive strategy for network slicing designed to bolster seamless communication and operational efficacy of emergency response teams in varied and ever-changing conditions. It presents an innovative approach to managing workload fluctuations in network slicing. Also, it introduces a new slice configuration mechanism to prioritize signals for devices within the complex, compelling, hierarchical network systems. Incorporating a Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm is central to the approach, tackling the complexity of implementing effective communication strategies across multiple network layers. Our findings demonstrate a highly adaptable and real-time slice configuration technique within System of Systems (SoS) environments, offering significant enhancements in systems engineering and emergency communication management. This approach contributes to the robustness and reliability of emergency response communications and underscores the importance of integrating environmental sustainability in developing next-generation network technologies.

Original languageEnglish
Pages (from-to)1752-1761
Number of pages10
JournalIEEE Transactions on Green Communications and Networking
Volume8
Issue number4
DOIs
Publication statusPublished - 2024

Keywords

  • Network slice
  • multi-agent deep deterministic policy gradient
  • multi-agent deep reinforcement learning
  • system of systems
  • virtual network functions

Fingerprint

Dive into the research topics of '6G+ Networks Through Enhanced Efficiency and Sustainability With MADDPG-Driven Network Slicing in SoS Environments'. Together they form a unique fingerprint.

Cite this