IEEE Wireless Communications · 2023 · 38 citations · 14 references
ComputingCloud ComputingComputer ScienceArtificial IntelligenceCloud ArchitectureEdge ComputingEdge ArchitectureEdge Artificial IntelligenceEdge NetworkingEdge Resource ManagementEdge-cloud ArchitectureEdge-cloud IntegrationCloud-native ComputingEdge Computing ArchitecturesEdge Computing NetworksEdge IntelligenceEdge-ai IntegrationEdge-cloud NetworkingMicroservices DeploymentZero-touch Provisioning
Network agility, automation, and intelligence are at the forefront of the next-generation networks (NGNs) vision, which aims to provide zero-touch service management and self-optimizing networks. In this article, we give an overview of the significance of artificial intelligence (Ali-enabled NGNs, their projected benefits, design requirements, and critical challenges for evolving heterogeneous softwarized networks where microservices can be autonomously orchestrated, scaled, and maintained. The convergence of emerging disruptive technologies, for example, AI, network softwarization, hybrid cloud/edge-native computing architecture, with NGNs accelerates the enhanced service-oriented architecture at the network core/edge level to support on-demand microservices, such as visibility services for intelligent network management. In addition, we present a use case study and conduct experiments based on a novel design of an edge intelligence framework that orchestrates and deploys AI microservices utilizing the testbed resources of a multisite cloud/edge-native NGNs. We use a deep learning-based forecaster model to predict near real-time edge network flow between a centralized service orchestrator hub and multiple edge devices, geographically apart. The obtained results show that the deployed forecaster model accurately predicts the throughput and latency of edge network flow (verified against the groundtruth observations), which is additionally validated through two performance metrics obtained, low root-mean-square error, and high coefficient of determination values. Finally, we outline some of the potential future prospects for AI-enabled NGNs research.
14
Optimizing Computation Offloading in Satellite-UAV-Served 6G IoT: A Deep Learning Approach
Bomin Mao, Fengxiao Tang, Yuichi Kawamoto et al. · IEEE Network · 2021 · 173 citations
Toward Self-Learning Edge Intelligence in 6G
Yong Xiao, Guangming Shi, Yingyu Li et al. · IEEE Communications Magazine · 2020 · 159 citations