5G networks will incorporate new innovative technologies and concepts such as network virtualization, SDN/NFV, multi-tenancy and network slicing. Moreover, resource orchestration play a pivotal role to dynamically deploy network services and allocate resources. Orchestration is a control function for resource management in the network core and centralized cloud infrastructure. However, for end-to-end slicing and resource management, the orchestration functions must also be realized at the network edge i.e., the RAN segment. In this paper, we present an architecture for active RAN resource orchestration in which radio resource allocation to different tenants is dynamically scaled in real-time. A parallel can be drawn to the classic spectrum sharing scenarios as we evaluate the well known co-primary sharing model in a multi-tenant RAN context. Moreover, we extend SimuLTE, a well-known system level simulation model to integrate our RAN orchestration architecture and implement different scheduling policies. We evaluate system level network performance using fine-grained radio resource sharing approach and evaluate the involved performance-fairness trade-offs.
RAN Orchestration: A New Approach to Spectrum Management in Multi-Tenant 5G Networks
Shah Nawaz Khan;Leonardo Goratti;Roberto Riggio
2018-01-01
Abstract
5G networks will incorporate new innovative technologies and concepts such as network virtualization, SDN/NFV, multi-tenancy and network slicing. Moreover, resource orchestration play a pivotal role to dynamically deploy network services and allocate resources. Orchestration is a control function for resource management in the network core and centralized cloud infrastructure. However, for end-to-end slicing and resource management, the orchestration functions must also be realized at the network edge i.e., the RAN segment. In this paper, we present an architecture for active RAN resource orchestration in which radio resource allocation to different tenants is dynamically scaled in real-time. A parallel can be drawn to the classic spectrum sharing scenarios as we evaluate the well known co-primary sharing model in a multi-tenant RAN context. Moreover, we extend SimuLTE, a well-known system level simulation model to integrate our RAN orchestration architecture and implement different scheduling policies. We evaluate system level network performance using fine-grained radio resource sharing approach and evaluate the involved performance-fairness trade-offs.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.