The IP Development Network Blog

I chaired yesterday's session of the IIR Local Loop Unbundling conference. The brief was "Understanding the strategic, regulatory and cost implications of LLU to maximise ROI and sieze the competitive advantage". Rather long winded perhaps, but you get the picture...

The day started with a presentation by Gavin Young, Technical Committee Chair of the DSL Forum, who gave us an overview of recent developments and how these support new service offerings. One of the key points that Gavin made was that ISPs should be deploying MSANs rather than DSLAM or Gateway technologies in order to future proof themselves, even if they are not today planning on offering voice using full MPF.

The other point that Gavin made and one I strongly agree with is that ISPs need to be using TR069 standards to enable remote activation and management of consumer CPE. We are moving into a world where the IP break out into the home is going to get an awful lot more complex - its no longer just about connecting the PC, but the TV and other devices too. Users cannot be expected to manage the Home LAN that this creates, so ISPs (if they want to offer such services), need to be appying the lessons learned from many years of offering managed services to business customers.

Consumer Managed Services are where we need to be as an industry and the only way to do this cost effectively is to enable service agents to bulk configure and remotely manage the customer's routing in the home. TR069 isthe gobal standard that enables this.

Then we had Karen Larsen, Attorney-at-Law at Kyed & Jybaek who quickly dealt with my opening question "do we really need regulation at all?" It was a provocative comment, but Karen's view was simple and effective: regulation is essential because one entity has assets that their competitors need. Its obvious, but its the simple things that make the most sense and this was as simple as it gets. Karen also discussed the issues that we face as an industry dealing with regulation and the problems with enforcement. When there is a breach, the typical response is to plead ignorance and rectify the behaviour - which is fine in isolation but is open to abuse...

The panel session included Gavin, Dana Pressman (MD, Be Unlimited), Sebastien Pham (Service Delivery Manager, Telecom Italia France) and Richard Morecroft (Head of LLU at The Carphone Warehouse). The opening remarks detailed the capex intensive nature of the business we are in, but how ISPs should future proof their networks by deploying Gigabit rather than 100 Megabit links to exchanges.

My view is that although sound technically, deploying gigabit links, ujnless you own the fibre (which very few do) is commerically is very hard to justify given the volumes that ISPs might expect to see from unbundled exchanges. With peak hour usage per subscriber at 30kbps, you need over 3,000 users on an exchange to generate more than 100Mbps worth of traffic. Even with traffic per user twice or three times that rate, you still need a lot of subs before you need even a second 100M link and given the cost of Gig links at ~3x 100M links you can start to appreciate that this may not be the most cost effective way to connect the customers to the core network. Multiply this cost by hundreds or thousands of exchanges and you have a very big bill for capacity that is not yet required. We will get there, but I would suggest deploying less initially and upgrading as requirements grow rather than carrying the extra cost while the subsriber base builds.

The panel also noted that LLU is a means to an end - very true, we are not selling copper pairs, are we? - and the end is that you are nearer and have more control over your customer.

I wanted the panel to think about whether the industry has a responsibility to provide universal access, not just high bandwidth to the cities, but to the towns and rural areas too. The consensus was that the industry or individual companies within it are not responsible and that the issue lies squarely with government (national / regional / local). I'm not sure about this though as in many countries the goverment does not own a telco any more - do we want a new BT just to provide coverage where it is commerially unviable? Or, if its subsidies, who gets the subsidy? The Regional Development Agencies in the UK tried to tackle the problem by creating their own public / private partnerships but although this provided startup capital, has been unable to sustain an ongoing business model.

There was a suggestion from the floor that rural users should pay more than those in the cities: the flip side is that car insurance costs more for city dwellers than those that live in the countryside so why shouldn't the reverse be true? This does make sense and could solve some of the problems, but may not be enough to deliver access to those on the smaller exchanges where costs may run into hundreds of pounds per customer just to get the infrastructure in place. Would I pay £200 install and £35 per month to connect my village dwelling...? Probably not.

Moving on, Dana Pressman gave us an insight into how Be built and sold their business. The key messages were to know where you want to get to and make sure that everything you do is specifically designed to get there. They had a very small team (starting with 5 and growing to 25 or so people) and are rightly proud of achieving what they did with so little fuss.

After lunch, Jon Pearce from Zyxel gave us his view on the various xDSL technologies that exist and what they can do. These are available today, but need very short loops and so we have seen deployments only in densely populated (mainly Asian) countries.

Sebastien Pham detailed how sub-loops are being unbundled in France, but gave a cost of €600 per customer vs "regular" DSL at around €150. No-one at the conference was quite sure how this cost can be paid back commercially and must put a lot of doubt around the sub-loop unbundling process which is being targeted at the 8% of the population where loops are just too long for anything that could reasonably be called broadband. Maybe this is the cost that governments need to fund?

I then had a slot and presented my slides which are a summary of research that I have done this year. My key point to ISPs is that they own the customer and need to look at finding ways to generate income streams from the content owners that so badly need the additional capacity that ISPs are deploying. Content owners are the ones getting paid by the consumer. ISPs are in a price death-match at the moment and need to find other sources of revenue from those who benefit most from their investments.

I also highlighted the case of Hanaro Telecom who are to my knowledge the first IPTV Station to have their traffic blocked by ISPs because of the volumes that their content has caused. This changes the game significantly. ISPs need to stand up for themselves and not be frightened of losing a few (very high usage) subscribers in order to protect their investments. So what if these users go to your competitors...? Wouldn't getting rid of power users actually be a good thing?

At least my position on Network Neutrality is clear...

We ended the day with a presentation from Richard Morecroft (CPW) on the different solutions to delivering voice over LLU networks. He clearly believes that Full MPF is the way forward as VoIP has a lot of issues around it as a straight replacement for POTS and the additional benefits are marginal and do not outweigh the downsides.

You can see his point: POTS isn't broken so why fix it?

I must admit to having previously held the view that VoIP was the way forward because it is fundametally more efficient. There are solutions to the major problems too (emergency service geo-positioning, QoS, installation and support), a lot of which involve the TR069 standard that Gavin alked about. Richard's comment was that it depended on where you were coming from and they have a large existing voice business that can be migrated to Full MPF much easier than to VoIP.

All in all, a very interesting day. We can see that LLU is now a very real, mass market product (should that be product enabler?). BT announced last week that over 1 million lines are now LLUd in the UK and there is determination from all sides to make it work. I just wonder how long it will be before they hit my little village in rural Cambridgeshire...?

# posted by Jeremy Penston @ 11/14/2006 03:37:00 PM

Tuesday, 14 November 2006

IIR LLU Conference