Slice-aware energy saving algorithm for 5G networks based on simplicial homology.

Network slicing is an important feature introduced by 5G. It enables an operator to deploy multiple independent end-to-end virtual networks, called network slices, on top of a common physical infrastructure. Network slices require the allocation of resources at different levels of the system, among which power budget is of upmost importance at RAN level. This work proposes a new power allocation algorithm for 5G network slicing. Network slices are introduced as a set of five parameters that reflect slices capacity, latency and reliability requirements. Simplicial homology is used to analyse network coverage. Reliability is then defined in terms of a multiple network connectivity criteria. The performances of the proposed algorithm are evaluated by simulations for a scenario with three different network slices: eMBB, uRLLC and mMTC. This method can be used to determine the required power budget at RAN level for different 5G network slices deployment scenarios.