Virtual resource allocation for wireless virtualization networks using market equilibrium theory

Abstract

This paper proposes a virtual resource (VR) allocation scheme for OFDMA based wireless virtualization networks (WVNs). The scheme considers the time-variant wireless channel quality from the technical perspective, as well as the network benefits for WVN elements, e.g., subscriber user equipments (SUEs), mobile network operators (MNOs) and wireless service providers (WSPs), from the economic perspective. Regarding the available VR for a WVN as commodities, the utilities of SUEs and the network profits of MNOs and of WSPs are defined by using logarithmic type utility functions. Considering each of the MNO and the WSPs is willing to maximize its own network profit through the VR sharing, the VR allocation between one unique MNO and multiple WSPs is formulated as a multi-objective optimization problem. To solve the problem, we first decouple it into two single-objective problems: the first is to maximize the benefit of the MNO at the supplier-level, and the second is to maximize the benefits of the WSPs at the consumer-level. By searching the equilibrium price of the VR market, the original multi-objective problem can be solved in the Pareto optimal sense. The convergence of the proposed heuristic algorithm and the effectiveness of the proposed VR allocation scheme are testified through extensive experiments.