An Efficient Model for Mobile Network Slice Embedding under Resource Uncertainty

The fifth generation (5G) of mobile networks will support several new use cases, like the Internet of Things (IoT), massive Machine Type Communication (mMTC) and Ultra-Reliable and Low Latency Communication (URLLC) as well as significant improvements of the conventional Mobile Broadband (MBB) use case. End-to-end network slicing is a key-feature of 5G since it allows to share and at the same time isolate resources between several different use cases as well as between tenants by providing logical network. The virtual separation of the network slices on a common end-to-end mobile network infrastructure enables an efficient usage of the underlying network resources and provides means for security and safety related isolation of the defined logical networks. A much-discussed challenge is the reuse or overbooking of resources guaranteed by contract. However, there is a consensus that over-provisioning of mobile communication bands is economically infeasible and a certain risk of network overload is acceptable for the majority of the 5G use cases. In this paper, an efficient model for mobile network slice embedding is presented which enables an informed decision on network slice admission. This is based on the guaranteed end-to-end mobile network resources that have to be provided on the one hand and the capacities and capabilities of the underlying network infrastructure on the other hand. The network slice embedding problem is solved in form of a Mixed Integer Linear Program with an uncertainty-aware objective function. Subsequently, the confidence in the availability of each resource is analyzed.