Friday 25 April 2008

Draft: Explanation of Peering and Transit

I've written an explanation of Peering and Transit. Which is available in PDF in draft. I would love to receive comments on it from anyone who is interested. My main questions are: Did I cover all angles, is it readable, is it structured well and does it need anything else? Please send your comments to the e-mail address in the paper.

You can find the paper here:

UPDATE: now available on Ars Technica

Tuesday 22 April 2008

Mobile Terminating Access tariffs to go to 1 cent in 2012

08-05-2007 The presentations of the Encore workshop are available. They all argue forcibly in favor of diminishing or abolishing MTA-tariffs. For those of you who want to know about Peering and Transit as the alternative model, see my draft explanation.

Reuters last week reported that the EU will dramatically cut Mobile Terminating Access to a level of 1-2 cent per minute. " "The logic of the recommendation would be that by 2012 the incentive for keeping fixed termination rates and mobile termination rates would go," the Commission official said." I had missed this, but today this was confirmed at a workshop organised by Encore in The Hague. An official working for one of the EU regulators said 1-1.5 cents would be the goal of the Commission.

There is a great deal of logic behind this number. Currently the termination rates vary from 2 cents in Cyprus to 16 cents in Italy. Many countries are now forcing rates down to about 6 or 6 cents. Which is still well above cost according to ARCEP who are of the opinion that current costs are 3-4 cents and 2.5 cents is what their models predict the cost will be in 2010. So cost can come down dramatically without dropping below costs in the current Long Range Incremental Costs models.

The Commission wants to be forceful and will therefore not shy away from a statement. It cannot state a number that regulators already envisage. The start for the Commission will be the current MTA-tariff of Cyprus. It will not follow the gradual glidepaths of regulators. It doesn't want 2.5 cents by 2020, but it will want it next year or the year after. Bij 2012 this will need to be at 1 cent, as this is where mobile to fixed and fixed-fixed MTA-tariffs are at. Anything higher is not a proper statement.

The Commission can do this because there is no proper controlling force of the Commission. A recommendation will work as a regulation in the work of most regulators. If the regulator comes to the Commission with a higher MTA-tariff, the Commission will refer to its recommendation and the MTA-tariff of 2.5 cents. It will then start the serious doubts procedure, causing all sorts of political trouble for the national regulator while making the Commission look all harmonizing and consumerfriendly. So this is a lose-lose situation for the national regulator.

The Commission may have won part of the roaming battle, but it didn't succeed fully, because it wanted a directive and therefore had to agree to some changes in order to get the support from Vodafone.... I mean the UK. By doing a Recommendation the Commission may save itself the trouble of listening to the operators (countries) protest in all kinds of EU Council meetings. The Commission will look benign and any regulator not following the guidelines will look incompetent.

I don't belief the Commission will force Bill and Keep on the market. It will hope that by bringing the MTA-tariff's to the same level as fixed-fixed the market itself will choose to abolish MTA's all together.

Friday 18 April 2008

Thursday 17 April 2008

Why no one may dig fiber to your home (OECD presentation)

Whether or not you will get fibre to your home is a tug of war between various actors. In this presentation for the OECD Fibre Workshop in Stavanger Norway, I tried to describe this tug of war between the players. Though I show that the government is the party with the money and the possibility to monetize the benefits of the network, this doesn't imply it should (and in many countries its doubtful it is legally allowed.) Who should and who can depends on where you are and the competitive situation there. See also the OECD Fibre paper mentioned further on this blog

Monday 14 April 2008

ARS Technica has two good stories on future of the net.

"Ars Technica had two good stories on the future of the Net. One story examines the future of broadband networks and how ferrets will be bringing it to you. The article points to the OECD-study on how we will need 50Mbit/s minimum in the coming years, which was mentioned earlier on the blog. (Also explaining why squirrels and sharks are natural enemies of the internet, which ARS didn't pick up.) I do find it amazing to see what the press does with a paper like the OECD future of fibre paper. Mostly they do nothing with it (because the OECD doesn't send out press releases) but also they pick up on one line in the entire paper. It is however a fun way of looking at things.
The other one is an interview with Andrew Odlyzko of the Minnesota Internet Traffic Studies (MINTS) project, explaining how the predicted Exaflood that would overflow the web is not happening and the growth of traffic is actually slowing and there is no sign that ISP's cannot keep up with bandwidth growth."

Friday 4 April 2008

Technology, economics, ethics, qos and flow control

Dr. Lawrence Roberts, who was one of the men who started the internet and is no head of switch company Anagran wrote on Internet Evolution. I reacted as follows:


I think I will have to disagree with the whole contents of your post. The problem lies in a combination of factors: technology, economics and ethics. Technology wise I do wonder about the technical soundness of flow control routers. We tried much of the basics with ATM and the idea didn't scale well. However, you've got your technical credentials, so lets assume the product you are pitching works as sales sais it does.

The problem however is that technology in itself may be sound, but that the discipline of economics provides the measures along which the success of technology gets measured. Terms like: the best, fairness, quality, simple and even scale are not technical terms. There is no best chip or router design. ATM isn't better or worse than Ethernet. It does different things. Simplicity is impossible to measure when you're talking millions of lines of code, high end lasers, asics etc.

Measuring what is best requires therefore insight into the field of economics. I've once tried to write an outlne for a scientific paper on the subject. You can find it here: http://lunaticthought.blogspot.com/2007/11/there-is-no-economic-basis-for-qos.html One of the main problems with adding flow control like structures to networks is that they base themselves on the premise that the network is almost full and therefore we need a mechanism to manage congestion on the network.

There are 2 problems with this idea:

1. Is there a problem: Are networks really full? (which means in technical terms that spikes at small time intervals (hundreds of miliseconds) are exceeding network capacity? Are we really seeing this often?

2. Are you the solution: Is the solution to managing a traffic jam, to introduce congestion management or to increase capacity? Network growth is 50% per year according to mr. Odlyzko. So when the current usage of the network is set at 100 and it is reaching the capacity where flow control kicks in.. In a year time it will be 150 and in two years time at 225. How many packets can flow control drop without it being like TCP?

With TCP fairness (fairness is ethics) ideas there are also similar problems.

1. It assumes that over the entire stretch between two network points there is a problem. What if the two end points are small say 1mbit upload on both sides. Why would that require control in a non congested 10Gbit network in the middle?

2. It seems to assume that every TCP flow should get an equal proportion of bandwidth. However I have paid for 20/1 DSL. What would be a fair distribution of available space on a 100mbit line between me and 40 people with a 5 mbit line? Do we all get equal amounts of bandwidth? Do I get 4 times as much because I have bought 4 times as much? or 8 times as much, because I pay 8 times as much. Do we want communism or capitalism?

3. You mentioned that TCP punishes those that are further away. This sounds correct from an economical point of view, distance incurs costs, costs decrease efficiency. Inefficient solutions should be thrown out. Get stuff locally.

4. randomly dropping packets is seen as unfair. many solutions are trying to think up ways to determine which packets should be dropped. This requires the network to know the utility function of all users in order to determine fairness. However if I have 2 streams running each consuming 5 mbit/s whereas there is only one 9 available, maybe I don't want 4.5mbit per stream, but want to drop one of the streams. How would the network know which one to drop. Random dropping hits everybody equally with the added costs of omniscience.

5. How do you balance streams fairly if you don't know the future? traffic changes on miliseconds. Are you sure that dropping this packets is necessary, maybe the end-user stops a transfer and other packets should be dropped from the que. or maybe no packets need to be dropped, because all of the sudden traffic stops.

All in all if we look at this from an economics point of view, the rules of economy teach us that generally the solution is that one which is the most efficient as seen from the market players individually and as a whole. It should therefore have low entry costs, low transaction costs, little overhead and generally not make assumptions on what individual actors want.

All of these considerations I miss in your proposal. Plain flat Ethernet/TCP/IP maybe dumb and stupid, but it's dumb and stupid for everyone. It's not pretentious. If it's full you upgrade capacity. If there is enough capacity. There is enough for everyone.

OECD publishes (my) paper on Fibre

The OECD has published a paper titled "DEVELOPMENTS IN FIBRE TECHNOLOGIES AND INVESTMENT". Technicaly it's an OECD paper and it represents what the OECD finds important, but I spent two months in Paris (dec 2006-jan2007) writing it, so emotionally it's my paper. It was really great that my previous employer assigned me to the OECD for that period. The people at the Directorate for Science, Technology and Industry - Information and Communications Policy are great people and I had great fun working with them. These are the same people who make the infamous OECD Broadband Statistics, which are every time cause for discussion in the countries that are ranked low.

The paper examines developments in broadband technology, going from submarine, wireless, to hybrid (BPL, xDSL and Docsis) and all-fibre networks like (xPON and Point to Point). (Submarine is included because it both shows where the technology will go and the difficulties businesses may face) It describes the relative differences and the effects this has on the way these networks can be used. In one of the figures it shows the difference between the bandwidth on a dedicated link (like VDSL) and a shared link (wimax or Docsis) available to one user and to twenty simultaneous users. The paper is already a year old, so LTE isn't included. Service-wise it works from the idea that users need at least 50Mbit/s downstream capacity in order to be able to freely choose services from a variety of service providers and to be able to run several services parallel.


It also discusses the business case there are for investment in fibre networks and the different positions between incumbents and new entrants. Furthermore there is an evaluation of the possibilities for alternative financing models. It also shows a financial model based on a model for the Dutch Ministry of Economic Affairs. This model developed by Arcadis in the Netherlands is based on real investments in The Netherlands and therefore is quite accurate. When applying it to other countries your mileage may vary. What it does show is that housing density matters as well as penetration. What is also clear is that there is not much room for multiple providers of physical infrastructure in the same region.

The last part goes into the role of the government as a stimulator, regulator and investor. Some of the things it argues is that first governments should have an idea of what they want to achieve. Second it argues that governments should take away any barriers to entry, but also must be aware that if they stimulate one party that this interferes with competition for others. So if governments stimulate the roll out of a network, they should demand that the network that they have stimulated becomes open to others under equal conditions.

With regards to regulation an important insight is that regulators should provide regulatory certainty with regards to the success of networks, not with regards to the competitive services offered over that network. Also policies should aim at the local exchange of traffic between networks, thereby relieving backhaul networks of unnescessary traffic.

As an investor governments should be cautious and determine the amount of market failure. If they do invest it should be limited starting with i and then if more is necessary moving up to iv.

i) Digging trenches and laying ducts, removing a significant part of the costs of rolling out a
network.
ii) Providing passive network infrastructure to which network providers can connect their active
infrastructure.
iii) Providing an active network over which others can provide their services.
iv) Providing services over the network to end-users.

All in all it was great fun writing this piece. Another paper was also published called: PUBLIC RIGHTS OF WAY FOR FIBRE DEPLOYMENT TO THE HOME There is a conference in Stavanger next week that will delve much deeper into this topic. I will be speaking there too.

Tuesday 1 April 2008

Operators down on Mobile terminating fees fear

Bloomberg is reporting that stocks in Vodafone, Telefonica, Portugal Telecom, and Deutsche Telekom dropped sharply yesterday. Morgan Stanley has written a report warning that the EU may sharply drop Mobile Termination Rates. If this would happen, than it would open up the market to price competition and would seriously reduce the income gathered from incoming calls from fixed networks. Morgan Stanley sees drops to €0.025, in line with a presentation from ARCEP.

In my opinion there is also a serious possibility that the EU will advice to equalize termination fees from mobile-mobile and fixed-mobile to the same level as mobile-fixed and fixed-fixed. For The Netherlands this may mean a reduction of 80% all the way down to 1 cent per minute. This may not be in the very short run, but in the longer run.

Mobile operators would be wise to oppose any cut in termination rates as it would seriously impact their business. Not only would less money flow in, but it would also mean that there is no force stabilizing prices in the sector anymore. No competitor will price under the termination rate at the moment as this might mean that the customer becomes a bleeder. When termination rates drop to 1 cent, operators may just offer unlimited calling plans, like in the fixed line world. This leads to price competition and the erosion of margins.

Regulators are seriously looking into providing symmetry between Fixed Terminating Rates and Mobile Terminating rates. The ERG has had a draft common position online suggesting as much. In India termination fees are already symmetrical between fixed and mobile networks and growth doesn't seem to be a problem with 8 million new subscribers per month.

One thing that amazed me is that there is no mention of abolishing termination fees altogether in the ERG draft common position. This sounds rather novel, but is common practice in the internet world, where it is known as peering and transit and also in the United States and Singapore. The European Commission has a report out on this subject written by WIK consulting. It would allow regulators to step back from the market and not have to regulate it anymore at all, just like the case is in the internet world.

One benefit of sticking to termination fees, but dropping them sharply by equalizing them is that it takes away the "paying for incoming calls"- argument by the mobile telco's. They have consistently argued that abolishing terminating rates would require them to charge the subscriber for both incoming and outgoing calls, something customers fear (justly with todays prices in the EU). With a termination fee of 1 cent this argument wouldn't hold as strong.

To find some info look at my previous rant and posting the subject. Also have a look at the Encore Workshop on April 22nd in Amsterdam on the subject, where Carlo Cambine, Stephen Littlechild and Scott Marcus will speak. I plan to attend the session too.