SDN-Based Resource Allocation for Multi-Channel Wireless Mesh Networks

Abstract

In emerging multi-radio multi-channel wireless mesh networks (WMNs), how to allocate network resources to provide individual users with their fair rate share is a central but very complex issue due to their inherent multi-channel diversity and multi-hop connectivity. In this paper, we attempt to apply and extend a software-defined networking (SDN) approach to user-level fair resource allocation problems where both proportional fairness and max-min fairness are examined. We first mathematically formulate fair user resource allocation problems in multi-channel WMNs by seeking the maximization of objective functions under the network utility maximization framework. We then design both SDN controller-side and user device-side algorithms to solve the formulated problem in a centralized holistic manner. We apply the algorithms to control per-user link-layer rates according to a fairness criterion while we resort to traditional transport-layer congestion control for the regulation of individual flows for each user. For performance evaluation, we conduct a system-level simulation study to verify the convergence property of the proposed user fairness resource allocation solution and highlight the benefits of the SDN approach.