On achieving fairness in the joint allocation of buffer and bandwidth resources: Principles and algorithms

Fairness in network traffic management can improve the isolation between traffic streams, offer a more predictable performance, eliminate certain kinds of transient bottlenecks and may serve as a critical component of a strategy to achieve certain guaranteed services such as delay bounds and minimum bandwidths. While fairness in bandwidth allocation over a shared link has been studied extensively, the desired eventual goal is overall fairness in the use of all the resources in the network. This paper is concerned with achieving fairness in the joint allocation of buffer and bandwidth resources. Although a large variety of buffer management strategies have been proposed in the research literature, a provably fair and practical algorithm based on a rigorously defined theoretical framework does not exist. In this paper, we describe such a framework and a new, provably fair, and practical strategy for the joint allocation of buffer and bandwidth resources using the max-min notion of fairness. Through simulation experiments using real gateway traffic and video traffic traces, we demonstrate the improved fairness of our strategy in comparison to several popular buffer management algorithms. Joint management of buffer and bandwidth resources involves both an entry policy into the buffer and an exit policy through the output link. Our study reveals that, even though algorithms such as WFQ and DRR that can serve as fair exit policies have received significantly more attention, a fair entry policy is more critical than a fair exit policy to the overall fairness goal when buffer resources are constrained.