Quality of service and sending-party-network-pays: An absurd connection


Posted on September 12, 2012  /  2 Comments

Mr Luigi Gambardella of ETNO responded to one of my tweets and asked me to relook at their proposal. I did (CWG-WCIT12/C-109 of 6 June 2012). On the face, it appears that they are concerned about broadband quality of service, a real problem that we have been working on since 2007.

But then they go off the rails. The solution to QoS is supposedly treaty-level language mandating that “Member States shall facilitate the development of international IP interconnections providing both best effort delivery and end to end quality of service delivery,” and that “Operating Agencies shall endeavour to provide sufficient telecommunications facilities to meet requirements of and demand for international telecommunication services. For this purpose, and to ensure an adequate return on investment in high bandwidth infrastructures, operating agencies shall negotiate commercial agreements to achieve a sustainable system of fair compensation for telecommunications services and, where appropriate, respecting the principle of sending party network pays.”

This does not make any sense. Of course, it is a good idea for governments (Member States of the ITU) through their regulatory authorities to pay attention to QoS. We’ve been engaged with regulatory agencies seeking to improve broadband in multiple countries. Why is there a need for treaty-level language (“Member states shall”) to compel governments to do what they are doing anyway? Some are slow; we know. But pressure from stakeholders and media is what will get them moving, not International Telecom Regulations.

And anyone who has got their hands dirty working on QoS will know that it’s a complex problem, not all attributable to international factors. In PNG, we found in a preliminary diagnostic study that quality was atrocious even when we were downloading a packet from within the domestic ISP network. In other countries, we’ve found problems within the country because the ISPs were not handing over traffic internally, which was something they could do.

We have never said that international was irrelevant. We are working with UNESCAP to ensure there is enough international backhaul capacity in Asia, including redundant capacity. Nowhere in this process that included research and consultation with stakeholders did sending-party-network-pays appear as the solution. This is the centerpiece of the ETNO proposal. This is all there is in the ETNO proposal.

All the dancing around with mandating governments to do this and operating agencies to do that ends up here: give money to access network operators for the data terminated on their networks. Even if that would mean reengineering the entire design of the Internet; setting up government customs points to enforce the rules; and depriving the poor in our countries who are just beginning to use the Internet of most of the attractive and free content.

So, speaking as someone who has worked on QoS since 1998 (voice QoS then; when I was Director General of telecommunications) and upon careful analysis of the ETNO official text, I say this proposal should be defeated. We need to do the hard work of engaging with multiple stakeholders to improve broadband quality of service; not use it as a smokescreen to extract money for people who don’t know how to adjust to dynamic market conditions.

2 Comments