Network slicing is a widely discussed technology for 5G broadcasting services, as it allows the operators to create end-to-end virtual slices across the 5G core (5GC) and the 5G radio access network (RAN). The joint network slicing is necessary for 5GC and 5G RAN in a bid to enhance the overall resource utilization and quality of service. Nevertheless, as the dimension of network resource and user demand increase, the computational complexity of slice status update goes up as well. To overcome this scalability problem, we propose a data driven approach to realize joint slice reconfiguration from 5GC to 5G RAN in real-time. Our design shortens significantly the time of slice status update by mapping the resource status to a two-dimensional vector graph. In addition, we also introduce a corresponding slice reconfiguration algorithm to balance the load of each slice in 5GC and 5G RAN. Simulation results justify that our proposed approach can greatly reduce the cost of slice status update and achieve desirable slice reconfiguration performance.

中文翻译：

---

从核心到RAN的数据驱动网络切片，以实现5G广播服务

网络切片是用于5G广播服务的广泛讨论的技术，因为它允许运营商在5G核心（5GC）和5G无线电接入网络（RAN）上创建端到端虚拟切片。5GC和5G RAN必须进行联合网络切片，以提高整体资源利用率和服务质量。然而，随着网络资源的规模和用户需求的增加，切片状态更新的计算复杂度也随之增加。为了克服此可扩展性问题，我们提出了一种数据驱动的方法，以实时实现从5GC到5G RAN的联合切片重新配置。通过将资源状态映射到二维矢量图，我们的设计大大缩短了切片状态更新的时间。此外，我们还介绍了一种相应的切片重新配置算法，以平衡5GC和5G RAN中每个切片的负载。仿真结果证明，我们提出的方法可以大大降低切片状态更新的成本，并获得理想的切片重新配置性能。