As part of the wider debate over net neutrality, network management practices of Internet Service Providers have become an issue of public concern. Several times starting in 2010, the Federal Communications Commission asked for public input on which network practices are reasonable forms of network management. Little attention has been paid to this issue within the academic networking community, and most Internet policy researchers have recommended a case-by-case analysis.

In contrast, in these papers we propose a framework for the classification of traffic management practices as reasonable or unreasonable. To build the framework, we focus both on the technical aspects of traffic management techniques and on the goals and practices of an ISP that uses these techniques. The framework classifies traffic management practices as reasonable or unreasonable on the basis of the technique used and on the basis of how and when the techniques are applied. We suggest that whether a traffic management practice is reasonable largely rests on the answers to four questions regarding the techniques and practices used. We consider examples of how these techniques are used by ISPs, and how the answers to these four questions collectively affeect the degree to which a traffic management practice is reasonable. Based on these questions, we propose a framework that classifies techniques as unreasonable if they are unreasonably anti-competitive, cause undue harm to consumers, or unreasonably impair free speech.

We propose that traffic management practices are reasonable if they are implemented at endpoints, are chosen by the user, are based on reasonable application provider payment, or involve providing QoS to traffic chosen by the user. We propose that traffic management practices implemented in transit nodes without user choice are unreasonable if they block applications, assign QoS based on source, destination or service provider, or assign QoS based on unreasonable application provider payment. We suggest that QoS based on the application can be more effectively implemented by allowing the user to determine the priority of his/her applications, and we suggest that any charges for QoS can be most effectively implemented by integrating them into subscriber contracts and into the Service Level Agreements between ISPs, rather than by charging application providers that are not subscribers.

This paper is intended for people with a background in communications policy:

How to Determine Whether a Traffic Management Practice is Reasonable (with A. Ghosh), Research Conference on Communication, Information and Internet Policy (TPRC), Arlington, Virginia, September 2009.

This paper is intended for people with a technical background in networking:

A Framework for Classification of Traffic Management Policies as Reasonable or Unreasonable (with A. Ghosh), ACM Transactions on Internet Technology, vol. 10 no. 3, October 2010, pp. 12:1-12:23.

Portions of this work were supported by NSF. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation or IEEE. This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. One print or electronic copy may be made for personal use only. Permission must be obtained from the copyright holder for systematic or multiple reproduction, distribution to multiple locations via electronic or other means, duplication of any material in these papers for a fee or for commercial purposes, modification of the content of these papers, reprinting or republishing of this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, and to reuse any copyrighted component of this work in other works.