Profit-oriented resource allocation using online scheduling in flexible heterogeneous networks

Abstract

In this paper, we discuss a generalized measurement-based adaptive scheduling framework for dynamic resource allocation in flexible heterogeneous networks, in order to ensure efficient service level performance under inherently variable traffic conditions. We formulate our generalized optimization model based on the notion of a “profit center” with an arbitrary number of service classes, nonlinear revenue and cost functions and general performance constraints. Subsequently, and under the assumption of a linear pricing model and average queue delay requirements, we develop a fast, low complexity algorithm for online dynamic resource allocation, and examine its properties. Finally, the proposed scheme is validated through an extensive simulation study.