A Base Station Agnostic Network Slicing Framework for 5G

Abstract

The major investment required to roll out new 5G networks has been partially overcome with the definition of a set of migration paths from current 4G deployments to full 5G NR standalone deployments. However, these migration paths, which reduce the time to market of 5G services and the initial investment, result in heterogeneous networks with higher complexity. In this article, we propose a base station agnostic framework for Network Slicing (NetSliC), a new virtualization layer that creates slices taking into account a set of specific constraints per BS. The proposed mechanism is able to effectively deal with these new network architectures by abstracting the complexity of BSs and slicing the network to guarantee the traffic requirements. It is shown that NetSliC can outperform the existing literature. For instance, NetSliC achieves around 67 percent average throughput gain and 16.6 percent baseband unit processing load reduction when compared to baseline scenarios.