Resource Allocation for Virtualized Wireless Networks with Backhaul Constraints

Abstract

In this letter, we study resource allocation for wireless virtualized networks considering both the backhaul capacity of the infrastructure provider (InP) and the users' quality-of-service (QoS) requirements. We focus on the profit gained by a mobile virtual network operator (MVNO), which is a middleman who buys physical resource from the InP, bundling them into virtual resources called slides before selling off the service providers. The objective of the MVNO is to maximize its profit while guaranteeing the backhaul constraint and users' QoS by jointly allocating the slices and the uplink transmit power to the users. To solve the formulated mixed integer non-convex problem, we propose a distributed solution framework based on Lagrangian relaxation to a find suboptimal decision about slice and transmit power allocations. We further propose a low-complexity solution based on the concept of a matching game that does not require any global information. Numerical results are provided to evaluate the performance of the proposed schemes.