The recently concluded Broadband World Forum (BBWF) 2013 in Amsterdam would have been just about pure hell for devoted fiber-to-the-home (FTTH) evangelists, given that DSL-acceleration technologies such as VDSL2 vectoring and G.Fast dominated both conference and exhibition-floor discussions.
Just a few years after many proclaimed the impending death of copper networks, those very same networks are squeezing out previously unimaginable speeds, and the high-speed broadband game has changed completely.
Several trends emerged both on the show floor and in the conference sessions.
Get ready for vectoring
Unsurprisingly, “vectoring” was the term heard most often at the conference, with both major fixed-line vendors – Alcatel-Lucent and Huawei – pushing their vectoring technologies extremely hard and with several major European operators on the verge of commercial vectoring deployments after completing trials with vendors.
Belgacom confirmed plans to launch vectoring in Q114, though Switzerland’s Swisscom and KPN of the Netherlands might beat the Belgian incumbent to it. KPN’s commitment is the most remarkable: Not so many years ago, the Dutch incumbent was committed to the “purest” – and usually most expensive – form of FTTH, point-to-point fiber.
Interestingly, operators are split on how they will use vectoring. Some – Deutsche Telekom, for example – are determined to use it to launch 100Mbps packages to take on cable MSOs head-on, while others see vectoring more as a means to offer additional capacity at overloaded areas of their networks.
The devil's in the detail - vectoring and G.Fast
Despite the abundance of vendor charts detailing trouble-free trials, we heard that at least three operators’ field tests of vectoring have not gone so swimmingly.
One complained of being able to generate only single-digit-percentage gains in bandwidth, and another found that large parts of its network were entirely unsuitable for the technology. The fundamental problem was the poor quality of the copper cables, a situation that is costly for operators to discover, let alone rectify, given that the vast majority of lines were buried decades ago.
Nor are all vectoring chipsets equal, apparently. At the interoperability pavilion, one demo revealed that three modems using different chipsets achieved about 100Mbps speeds, while another reached 130Mbps.
Not quite up there with vectoring in terms of buzz but not that far behind it was G.Fast. This (potential) pinnacle of DSL-acceleration technology is clearly viewed by operators around the world as their most realistic means of offering mass-market 1Gbps broadband in the medium term, given the difficulties of deploying full FTTH.
Both Alcatel-Lucent and Huawei were pushing hard on their new fiber-to-the-distribution-point network technologies, with the former introducing its “micro-nodes” at the event. Both vendors say G.Fast will play a major role in the next three to five years.
A detailed model presented by Dutch non-profit consulting firm TNO suggested that the capex required to build a hybrid fiber-to-the-x and G.Fast network covering part of Amsterdam would be considerably less than that required for a full FTTH network, mainly because of the savings made by removing the need to lay extra fiber in streets and within homes. Yet neither TNO representatives nor anyone else Informa spoke to at BBWF was clear on what the opex costs of rolling out such a network would be.
The costs could be considerable. Whereas FTTH networks are relatively simple, with active network equipment sitting in just thousands of local telephone exchanges or central offices, G.Fast networks will require managing active equipment in tens – and perhaps hundreds – of thousands of public locations, such as telephone poles, manholes and the sides of houses.
One operator executive also argued that the capex case is far from made. He argued that given the volume of network equipment that an operator will need to invest in, vendors need to do a lot more to bring chipset prices closer to those of customer-premises equipment, rather than the considerably higher rates commanded by network equipment.
Although chipset vendors no doubt welcome the scale promised by G.Fast, they are likely to be less enthusiastic about the pressure it might put on their margins. Agreement between operators and vendors on a reasonable balance will be key to making the G.Fast business case work for all stakeholders.
For these reasons and others, many European operators are still on the fence about G.Fast. In any case, the pragmatism of using a variety of FTTx architectures depending on how the costs stack up means the technology will probably become just one option among many, even once the capex and opex costs are clearer.
Telcos to embrace....cable?
Deploying FTTH in apartment blocks and other multi-dwelling units (MDUs) can be a nightmare, and even using in-building copper can be problematic. It should come as no surprise, then, that vendors – and even some leading FTTx operators – said at the event that they are looking at using in-building coaxial-cable TV wiring to deliver high-speed broadband.
The largest opportunity appears to be in China, where 20-25 million homes are suited to be served by the technology, according to vendor estimates. But operators in Europe are also interested, and the initial market there could be as large as 1.5-2 million homes, according to vendors.
Three main technologies have emerged: a G.Fast-based technology being developed by Huawei; chipset vendor Broadcom’s C-DOCSIS system, which uses the DOCSIS cable-broadband technology as its foundation; and Swedish company InCoax’s technology, which is based on the MoCA standard for distributing TV and other media around the home.
InCoax says it has already received interest from nearly half a dozen European telcos and strong interest from operators in China, all of which are eager to see whether existing in-building coaxial cable can offer them a more efficient high-speed-broadband solution.
Using coaxial networks makes great sense from an operator perspective because it embraces the philosophy of using existing infrastructure rather than building expensive new networks. One major stumbling block could be the ownership of in-building cable networks. In Western markets, cable operators will probably jealously guard theirs, and in many others, ownership is complicated or unclear. Who owns the network, for example: the operator, the building owner, the housing association or the local authority?
Two of the most interesting conversations I had at the conference were with Asia Pacific operators – both from countries viewed as “heartland” FTTH markets – who commented that they were keeping a close eye on DSL acceleration and in-building coaxial technologies.
Both operators said that they have huge numbers of MDUs in their network footprints, which they simply cannot afford to rewire for FTTH. They are looking at ways to offer increased speeds in a cost-effective manner, and using G.Fast or even souped-up coaxial cable is on their menus.
Google Fiber keeping the dream alive
One of the most highly anticipated keynotes from the event came from Google Fiber general manager Kevin Lo, although, truth be told, his presentation contained more froth than substance. Many delegates were left with plenty of questions as to what Google is really up to in the access business.
Unsurprisingly, Lo insisted that there really is a market for gigabit Internet and again stated his belief in the “if you build it, they will come” theory that applications to fill gigabit connections will arrive once users have access to the bandwidth.
However, Lo, who took only two questions in his Q&A session – both of which were frivolous – did not get down to the central question that everyone wants answered: Does Google really want to be a mass-market network operator?
Google Fiber certainly seems to be working as a lobbying effort. The flimsy – or perhaps nonexistent – evidence Lo presented was seized on by certain media outlets and evangelists as “proof” of the need for ultrafast broadband now.
Frustrations of the gigabit frenzy
Although Lo talked up the potential for gigabit broadband, plenty of other operators – including an extremely high-profile operator from a cutting-edge FTTH market – complained about the frenzy for nationwide gigabit broadband, saying that it is unnecessary and unaffordable for operators at this point.
“We are already offering our subscribers average download speeds of 70Mbps,” an exasperated executive at the high-profile operator told me. “There is nothing they cannot do with their current speeds. Yet all the time we are getting pressured by the government to upgrade everyone to gigabit speeds so that we can look like a technology leader – it is very frustrating.”
Another very interesting conversation at the conference – with another executive from a leading Asia Pacific telco – centred on the fact that operators are beginning to realise the limits of their “salami” strategies to add more meat to their broadband-access products in the form of value-added services.
The operator in question has launched an extraordinary range of value-added services over the last couple of years and last year made an aggressive leap into the education-provision market. But the executive says his company has learned the hard way that entering new sectors is hard work.
“We thought we could enter the education market with our own services, but it has been a lot harder than we thought,” he said. “There are a lot of strong players already in place, and we are now adjusting our strategy to try and bring more partners on board.”
FTTH has to look good
One of the most interesting vendor announcements at the conference came from Swiss company ADB, which announced the launch of its new – and vastly simplified – in-home socket for GPON-based FTTH networks.
ADB officials say that they launched the product – little larger than a domestic electric plug – after noticing the significant percentage of subscribers who cancelled FTTH installation when they realized engineers would have to install several unsightly grey boxes on their interior walls.
The ADB GPON optical-network terminal (ONT) is a small, unobtrusive, white device that links to a home gateway via Ethernet, which even the fussiest homeowner or landlord would be hard-pressed to object to.
How to ensure that operators can take advantage of devices such as ABD’s ONT was confirmed as a key problem for the industry to solve.
In conjunction with the Broadband Forum standards body, Informa shared the results of an extensive study, “Optimizing the fiber business case: GPON interoperability strategies for operators and vendors”, examining the opportunities and challenges related to improving the potential for GPON equipment from different vendors to work together.
Tony Brown is a senior analyst with Informa Telecoms & Media. He is a key member of the Broadband and Internet Intelligence Centre team, covering the broadband and Internet markets of the Asia Pacific region.