A Federated Deep Reinforcement Learning Approach for Distributed Network Slicing Orchestration
Rezazadeh, FarhadDevotiy, FrancescoZanzi, LanfrancoChergui, HatimCosta-Pérez, Xavier
Abstract
Enabled by the network slicing technology, multiple and independent virtual networks can now be instantiated and customized to meet heterogeneous service requirements over 5G network deployments. The widely different requirements of 5G emerging use-cases such as Internet of things (IoT), augmented/virtual reality (AR/VR), vehicle-to-everything (V2X)communication, exacerbate the need for orchestration solutions able to concurrently accommodate these services in a resource and cost-efficient manner. Nowadays, centralized orchestration solutions are available. Such solutions require a holistic view over multiple networking domains, thus suffering severe scalability limitations that hardly fits with latency and efficiency requirements of 5G deployments.
Metrics
Topics
Related Documents
Network Slicing via Transfer Learning aided Distributed Deep Reinforcement Learning
Tianlun HuQi LiaoQiang LiuGeorg Carle
Using Distributed Reinforcement Learning for Resource Orchestration in a Network Slicing Scenario
Federico MasonGianfranco NencioniAndréa Zanella
Two-Tier Resource Allocation for Multitenant Network Slicing: A Federated Deep Reinforcement Learning Approach
Ruijie OuGuolin SunDaniel Ayepah-MensahGordon Owusu BoatengGuisong Liu
A Multi-Agent Reinforcement Learning Architecture for Network Slicing Orchestration
Federico MasonGianfranco NencioniAndréa Zanella