Competition between service providers with strategic resource allocation: application to network slicing

Abstract

We propose and analyze a business model for 5G operators. Each operator is entitled to a share of a network operated by an Infrastructure Provider (InP) and use network slicing mechanisms to request network resources as needed for service provision. The network operators become Network Slice Tenants (NSTs). The InP performs the resource allocation based on a vector of weights chosen strategically by each NST. The weights distribute the NST's share of resources between its subscribers in each cell. We propose a strategy profile in which the NST chooses weights equal to the product of its share by the ratio between the total number of subscribers in the cell and the total number of subscribers in the network. We characterize the proposed solution in terms of subscription ratios and fractions of subscribers, for different cell capacities and user sensitivities. The proposed solution provides the exact values for the Nash equilibrium if the cells are homogeneous in terms of normalized capacity, which is a measure of the total amount of resources available in the cell. Otherwise, if the cells are heterogeneous, it provides an accurate approximation. We quantify the deviation from the equilibrium and conclude that it is highly accurate.