ABS: Adaptive buffer sizing for heterogeneous networks

Abstract

Most existing criteria [3], [5], [9] for sizing router buffers rely on explicit formulation of the relationship between buffer size and characteristics of Internet traffic. However, this is a non-trivial, if not impossible, task given that the number of flows, their individual RTTs, and congestion control methods, as well as flow responsiveness, are unknown. In this paper, we undertake a completely different approach that uses controltheoretic buffer-size tuning in response to traffic dynamics. Motivated by the monotonic relationship between buffer size and loss rate and utilization, we design a mechanism called Adaptive Buffer Sizing (ABS), which is composed of two Integral controllers for dynamic buffer adjustment and two gradient-based components for intelligent parameter training. We demonstrate via ns2 simulations that ABS successfully stabilizes the buffer size at its minimum value under given constraints, scales to a wide spectrum of flow populations and link capacities, exhibits fast convergence rate and stable dynamics in various network settings, and is robust to load changes and generic Internet traffic (including FTP, HTTP, and non-TCP flows). All of these demonstrate that ABS is a promising mechanism for tomorrow's router infrastructure and may be of significant interest for the ongoing collaborative research and development efforts (e.g., GENI and FIND) in reinventing the Internet.

Appropriate buffer sizes for Internet nodal devices: A networks' topologies–based solution

“…A common question that is asked in this context is as follows: What will be the situation if the link to the input interface of a router is saturated by several flows (>8, since the BDP rule was obtained with a link that was carrying eight flows) and the flows' RTTs are different? Vu‐Brugier et al and Zhang and Loguinov have questioned the credibility of these buffer sizing formulas. According to these researchers, the formulas specify fixed buffer sizes, when in reality, network traffics are time– varying in nature.…”

“…This point of view inherently implies that buffers sizes should depend somehow on the topologies of the networks. Zhang and Loguinov have also contended that routers buffers' sizes are closely linked to the following critical performancemetrics: packet loss rates, end‐to‐end delay, and link utilization.…”

Buffer Sizing for 802.11-Based Networks

“…However, this approach is not applicable to WLANs since it requires a priori knowledge of the link capacity or line rate, which in WLANs is time-varying and load dependent. [34] introduces another adaptive buffer sizing algorithm based on control theory for Internet core routers. [24], [14] consider the role of the output/input capacity ratio at a network link in determining the required buffer size.…”

Perspectives on router buffer sizing

“…[23] also presents an adaptive buffer sizing algorithm (ABS) where the router adapts its buffer size to suit the dynamics of incoming traffic. Using the monotonic relationship between buffer size, link utilisation, loss rate and queueing delay, ABS aims to maintain system performance above a certain given target objective.…”