Slice-aware Capacity Management in Multi-Tenant Networks
09 December 2019
5G is envisioned to provide substantial increase in the network performance both in terms of achieved rate and delay. However, in order to satisfy these goals, the network operators are forced to make significant investments on the network infrastructure. Consequently a key question is how to invest the scarce economical resources to achieve maximum impact. Unlike the predecessor technologies, 5G contains a vast amount of services with unique requirements and priorities. From a network operator's perspective, solely focusing on the particular services makes it harder to determine the business value of a region. Therefore, in order to efficiently use the available expansion budget, the conventional approach of network dimensioning has be revisited. In this work, we propose a novel utility function based capacity management framework that can be automatized to dynamically and efficiently manage the infrastructure resources. Build upon three decision metrics, i.e. the urgency of the required resources, available investment budget and the expected impact of the capacity increase, our algorithm determines regions to invest in. Moreover, our analysis shows that the efficiency of the proposed framework does not strictly rely on the long term observation of the traffic demand.