Piers Daniell

BT announced with great fanfare the launch of Annex-M services on their ADSL2+ (21CN) platform allowing up to 2.5 Mb/s upstream (from standard 1.3 Mb/s). Whooohooo. Excuse me if you detect a tiny bit of sarcasm but really how can BT talk about this being new technology when we have been offering it for over four years? We interconnect directly into the BE network (wholly owned by Telefonica O2 group) and have been delivering up to 2.5 Mb/s upstream to customers for all of this time, even using Cisco hardware.

I mention Cisco hardware because most businesses like to buy it. There is an old saying in the industry that goes ‘you don’t get fired for buying Cisco’. Granted it is pretty expensive but in all the years of doing it I can count the failures customer have had on one hand rather than the calculator I would need to tot up other ‘cheaper’ manufacturer’s hardware.

So you would think that if you were building a national network and launched a new variant (ie Annex-M) you would want your technology to work with Cisco hardware so you could conquer the business market? Well believe it or not, BT have followed the route of Tiscali and ended up with a network that we can’t make work with Cisco hardware! We are struggling to make it work at all to be honest, and that is with a not insubstantial client base already using Annex-M on the BE network which we have been supporting for years.

If this is the best they can do, I assume we are going to have to wait for the fibre to the cabinet services before we start seeing anything near impressive. Or, most likely, someone else will come and do that first so BT once again follow rather than lead.

Interesting piece on thinkbroadband.com with regards to router speeds over the O2/Be network we did a few months back. Basically as part of the wholesale offering for ADSL2+ we had to collate data on all the types of routers they may use and test them under lab conditions.

Although the Cisco’s came out near the bottom they did perform well with upload and in my mind have proved the most reliable and still what I would recommend to customers.

As 2010 rolls in, I am reminded that a decade is a significant period of time and a good one to reflect on, especially in the world of IT and Telecoms. What have we achieved in the past ten years, and what, if any, predictions can be made about the next ten years? Well, for one, Fluidata couldn’t exist until 2004 on the back of ADSL technology, to launch as its own business until early 2006. Which means that the last decade has seen great advances in internet connectivity speeds and brought the word ‘broadband’ into common parlance.

We might be far behind many other countries in this race, but don’t forget in 2000 dialup was the norm and businesses like Freeserve dominated internet access with products offering up to 56 Kb/s. Now people have the choice of ADSL2+ offering up to 24 Mb/s (24,000 Kb/s in old money) or even 50 Mb/s with certain cable broadband offerings. Fluidata, for one, became a leader in bonding technology providing services over multiple carrier networks at speeds up to 60 Mb/s down and 8 Mb/s up with PureFluid. Even gigabit (1,000,000 Kb/s) speed is achievable with the use of fibre, as well as more affordable especially in urban areas.

Networks not only got faster, but also more intelligent, with quality of service, large wide area private networks and MPLS protocols. Businesses put voice over the internet and now video is seeing a resurgence with advances in HD video conferencing. Which begs the question - what will happen in the next ten years?

With the industry talking about FTTH (Fibre to the Home) and FTTC (Fibre to the Cabinet) more people will be able to experience the high speeds that fibre can deliver. But remember this isn’t like the previous decade when the main infrastructure (the cables) remained the same. This technology requires something new which not only will absorb huge amounts of money, but also time and resource. The cynic in me would say BT’s 21CN (to provide ADSL2+ and more IP services) offering is aimed purely at the LLU carriers such as Tiscali, Carphone Warehouse, Cable & Wireless and O2. I am sure its reach won’t extend over 2,000 exchanges by the end of the next decade. So FTTH or FTTC products are purely to take the fight to the cable operators such as Virgin. And what will quench our thirst for data? I believe there will still be some advances in DSL with products such as VDSL, but also other completely different technologies which will see in excess of 100 Mb/s over copper. Bonding will become more prevalent and give rise to a larger number of niche ISPs able to aggregate multiple networks together, not only to improve performance, but also reliability. Fibre will of course continue to grow in popularity and while still not being within easy reach of home users, will hit more of the SME market as costs reduce.

Other technologies such as WiMAX will become more widespread as companies such as Intel start to bundle the protocols into their mobile chipsets. I know of one company soon to come to market with a 3 GHz service which, if widely adopted, could bridge the gap between broadband and 3G mobile networks. Another product I am looking forward to seeing more of is BPL (Broadband Power Line), which has the potential to deliver many hundreds of Mb/s anywhere in the country over a normal power line. It has been a long time in the making, but I believe this decade will see its widespread adoption.

Mobile broadband still has a long way to go so that future services, such as cloud computing, can be widely used. And what else is in store? Well to be honest, it can be anybody’s guess. The great thing about this industry is the fact it continues to innovate and change - I’m, for one, looking forward to the next decade.

There is a lot of talk at the moment about FTTH (Fibre to the Home) and FTTC (Fibre to the Cabinet) replacing DSL in the UK. However I think we are looking at many years before these technologies make any significant impact on the market as a whole. For one BT is protecting its business customers as much as possible, so the product is very much consumer oriented with high downloads, but crucially low upload speeds. Granted 2 Mb/s is a good upload for a customer but that speed is already served nationwide through the plethora of ADSL2+ exchanges with the likes of Carphone Warehouse and O2.

My feeling is that BT will continue to lower the cost of fibre to businesses, who still put a lot of value in this technology, to help drive access so that it can serve consumers in the long term. I have previously talked about their trials with FTTC with the issues surrounding the technology and scale needed for wide adoption of this technology.

So the short and medium term DSL still has a very important part to play in internet delivery and more so as businesses look to increase their reliance with the adoption of cloud and voice applications. We have been bonding DSL products for many years and with services capable of up to 80 Mb/s down and 10 Mb/s up there is a compelling case that FTTC and FTTH have a way to go.

We have been selling our BURST service for a number of years offering clients up to 20 Mb/s downstream and 2.6 Mb/s upstream across the UK utilising Telefonica’s ADSL2+ Annex-M network. Our recent announcement that we have lowered the contention guarantee to 5:1 further enhances this product offering. The VOX 2.0 product, which is delivered over the same network, offering (confusingly) up to 2.6 Mb/s symmetrically has had a reduction to 1:1 pitching it directly against IP Frame Relay and SDSL products.

So now businesses can be further assured of the high availability of these products and continue to benefit from the short contract term of 3-months, install time of a day and no download restrictions. It also means that bonding these lines together using PureFluid puts it directly in the path of leased lines delivering up to 60 Mb/s down and 8 Mb/s up.

With talk of people growing their own vegetables, using grease proof paper instead of plastic and not having any lights on to save the planet it looks like the internet will take the same twist. There was a funny piece in the news a few weeks ago about a chap in South Africa racing his pigeon against the speed of his internet connection.

Obviously the pigeon won, not least as he tried to send 4 GB of data over 60 miles using an ADSL connection. However in the UK that race would have been a bit closer on our ADSL2+ product. In our office at over 1,500 meters a line achieves 2.2 Mb/s upload which means it would take 5 hours to move 4 GB of data. However using our Gb/s internet connection it would have taken 40 seconds so the pigeon would have definitely lost out!

A lot of focus on the wholesale platform is giving other ISPs the ability to bond multiple lines together to get EFM speeds using the nationwide Be ADSL2+ network. We have had a lot of success recently with Cisco’s MLPPP protocol enabling multiple lines to act as one end to end. As you can see from the below graph showing a thinkbroadband.com speed test that it achieved 26.7 Mb/s download and 4.0 Mb/s upload on two lines bonded at over 1,500 meters from the exchange. This accurately doubles the capacity of the individual lines.

We are continuing to work with our partners on developing this offering to include more than two lines so that leased line speeds are achieved across this network.

We have some advertising running in the press at the moment on the new wholesale platform in partnership with Be. The platform allows for other network operators to have direct L2TP (Layer 2) access to the Be ADSL2+ cloud so that they can offer this network to their client base. Being the only network in the UK to fully support Annex-M (ie 2.6 Mb/s upload) and Cisco hardware there has been a lot of interest from ISPs looking to take this service on.

Another major milestone has been our ability to demonstrate end to end line bonding using Cisco’s MLPPP (Multilink PPP) protocol on this platform which partners can offer to their client. Being able to bond four lines to deliver up to 80 Mb/s down and 10 Mb/s upstream is a powerful message.

Lots of talk in the press recently on EFM (Ethernet First Mile) and thought it was worth a comment. What gets me is the fact that so many businesses in our industry see it as a new wonder product that will solve customers’ problems and deliver super fast connectivity over the existing copper infrastructure. From my perspective however it doesn’t seem to stretch the technology far enough to make the investment and move to it worthwhile.

Although the premise behind EFM is to bring ethernet protocols down to each site, the actual delivery in copper form, is still SDSL (albeit the .bis variant) and VDSL technologies. Therefore the product is still reliant on multiple copper pairs and very much subject to line length from the exchange or street cabinet. So from a technical perspective clients can get very much the same solution already using the PureFluid platform. The system is like EFM in that it bonds multiple copper pairs but differently it offers true carrier resilience as each copper pair can be provided by a different carrier. And where EFM will be built on one type of technology such as VDSL, PureFluid can bond different types of lines together such as ADSL2+ and SDSL delivering unique benefits.

However the biggest benefit is the technology is tried and tested and available nationally. With this, service can be achieved at 60 Mb/s making all but the very fastest EFM products redundant.

News from Thinkbroadband on Andrews & Arnold’s (AAISP) recent success installing FTTC into one of their client’s premises. Just goes to show BT are going to really shift their focus over the coming year towards fibre and away from their failing 21CN project. From AA’s recent issues with BT in providing their clients with a reliable 21CN they are obviously keen to find an alternative technology.

Essentially the FTTC is putting VDSL equipment into the local cabinet (instead of the exchange) dramatically lowering the distance from the DSLAM to the clients office. This ensures the speeds are better but also means that BT doesn’t have to go through the costly exercise of getting fibre into every building. This way as BT upgrade their existing 80,000 or so cabinets they can look to run fibre to them and hence offer VDSL services. I believe this is their preferred route to market as it would jump all the existing LLU carriers who have been deploying ADSL2+ for quite a while now.

However on first inspection it seems like a lot of work for not much gain. In this example the customer reportedly received 40 Mb/s down but was impressed by the upload of 1.7 Mb/s. Agreed that this is a big increase on what BT can currently deliver using their small ADSL2+ footprint but nothing compared to the 2.6 or even 2.7 Mb/s we have achieved with some of our clients using ADSL2+ Annex-M. Admittedly this is using the Telefonica O2 network, who we have an interconnect with, but still not the giant leap BT were hoping for. Also by installing two lines customers can receive over 40 Mb/s down with over 5 Mb/s upload as one service taking away the current gains of FTTC.

Obviously over time the footprint and technology will grow in very much the same way ADSL did when it first arrived in the UK. However I think we are looking at 18-months or so before you see any wide adoption of this promising technology.