Nº 2 2015 > New standard on G.fast
Get to Fast Faster, with G.fast
Stefaan van Hastel
Marketing Director for Fixed Networks, Alcatel-Lucent
Now that large-scale VDSL2 (Very high-speed Digital Subscriber Line 2) vectoring deployments are going live, G.fast is ready to take FTTx (Fibre-To-The-x) deployments to the next level. In 2011, VDSL2 vectoring put copper back on the map as a next-generation access technology by reaching the “magic number” of 100 Mbps downstream. Today, G.fast delivers hundreds of megabits per second — and Gigabit speeds are next.
By offering fibre-like speeds over existing copper telephone lines, G.fast allows operators to terminate fibre as close as possible to end-users, while leveraging existing telephone lines for the final metres or yards inside buildings. This way, technicians can avoid entering the building altogether — which is often a time-consuming and expensive part of any Fibre-To-The-Home (FTTH) roll-out.
Accelerating fibre deployments
Contrary as it might sound, G.fast will actually accelerate fibre deployments. “Fiberizing” an entire nation takes years or decades: workers have to enter every building to install new fibre cabling, which may also mean digging up every street and every front yard. However, most end-users are unwilling to wait that long, and policy-makers have set aggressive broadband goals. Operators need to get fast, faster.
G.fast provides an effective solution in places where installing new fibre infrastructure is uneconomical or infeasible. For example, older buildings are unlikely to come equipped with cable ducts, and some home-owners may refuse drilling or new cabling. Even in the best case scenario, the lengthy process of getting permission from the building owner, making an appointment, and entering the building is cumbersome, time-consuming, and costly.
G.fast avoids these complications by leveraging the existing, in-building telephone lines. Operators now have more options when seeking to deliver fibre speeds to end-users. Combining FTTH and FTTx deployment models is generally the most cost-efficient and fastest way to deliver ultra-broadband to end-users — effectively accelerating fibre roll-outs.
Bringing fibre to the most economical point
Where ubiquitous ultra-broadband is concerned, there is no single “best” solution, although FTTH is often identified as the most future-proof solution and end-goal for many operators. However, FTTx deployments often play a critical role in any operator’s strategy, due to its lower cost and faster roll-out compared with FTTH.
In FTTx deployments, fibre is terminated close to the end-user (the “x” can a node, curb, building, pole, manhole, wall, front door, etc.). From this “distribution point”, a fibre-fed G.fast (or VDSL2 vectoring) system delivers ultra-high speeds over the telephone lines. The trade-off is between speed and distance: G.fast uses a much wider frequency spectrum than VDSL2 to achieve higher speeds, but the maximum distance is reduced due to higher attenuation, down to typically 250 metres or less.
Obviously, shorter distances also mean higher cost, since fibre must be brought closer to the end-user. And since fewer end-users will be in range of the distribution point, a typical G.fast system will serve fewer end-users. An FTTN (Fibre-to-the-Node or cabinet) deployment with VDSL2 vectoring can deliver 100 Mbps downstream over 400 m, serving hundreds of subscribers. A G.fast deployment can deliver 500 Mbps+ aggregate over 100 m, but typically serving tens of subscribers in a building (FTTB), for example. As a result, FTTB/G.fast works about twice as expensive per end-user than FTTN/VDSL2 vectoring — but still about 30% cheaper than FTTH.
It is clear that operators need to choose between trade-offs between deployment speed, bandwidth, and cost. It might sound complicated and daunting at first, but this toolkit of deployment models (from FTTN to FTTH, and everything in between) allows operators to select the right tool for the job. For every city, street, or building, operators examine various factors (including the expected customer take-up rate; the availability of ducts; the type of soil, and whether it is easy or hard to dig; time needed to get permission; etc.) and select an appropriate deployment model that will connect these end-users in the fastest and most cost-effective way.
G.fast delivers on its promise
Since G.fast provides such a natural complement to fibre deployments, there is very strong interest from operators. Alcatel-Lucent has completed over 30 operator trials (by the end of the first quarter of 2015), and it is increasingly clear that G.fast delivers on its promise.
Figure 1 shows results from Alcatel-Lucent’s first series of operator trials, based on the first phase of the G.fast standard (up to 106 MHz of spectrum). The aggregate bit-rates are both high and consistent. At 100 m, 600 Mpbs or more can be achieved (the ITU–T G.9701 standard had originally targeted 500 Mbps). At 200 m, the trials show speeds of around 500 Mbps (the ITU–T G.9701 standard targeted 200 Mbps). Note that, contrary to VDSL2, all G.fast speeds are aggregate, i.e. upstream and downstream combined. Operators can choose how to allocate this aggregate bandwidth for both upstream and downstream, for example, through their management software — a major advantage.
It is important to note that all these tests used the VDSL2 spectrum. In networks where VDSL2 has already been deployed in the same binder, G.fast would be configured to skip the VDSL2 frequency spectrum (to avoid interference), resulting in bit-rates approximately 150 Mbps lower than the results shown here. The results above are based on the first phase of the ITU–T G.9701 (G.fast) standard, which provides up to 106 MHz of spectrum.
Typical G.fast deployment models
Fibre-to-the-Distribution Point (FTTdp) is the term generally used to indicate a G.fast deployment model. The distribution point can be any point close enough to end-users to accommodate typical G.fast loop lengths, and with access to the existing telephone lines.
Operators are interested in two types of G.fast deployment models: single-user and multi-user. In the single-user model, fibre is brought almost to the home (think FTTWall or FTTFrontDoor), and a single user is served from the G.fast node. In the multi-user model, multiple users are served from a single G.fast system (such as FTTBuilding or FTTManhole). In the multi-user model, neighbouring lines will interfere with each other, resulting in crosstalk and seriously degraded performance (speeds can be reduced by 50–90%). In order to enable multi-user deployments, the G.fast standard includes advanced vectoring capabilities to cancel this crosstalk and deliver near-optimal performance on each line.
Another practical but important consideration for G.fast deployments is powering. Due to the short loop lengths, deploying G.fast inevitably means rolling out tens or even hundreds of thousands of G.fast systems in the field. These systems need to be powered, and due to the number of nodes and their location (up poles, down manholes, or on walls), it is often not possible to rely on local AC (Alternating Current) power. Instead, a typical G.fast system can be powered either remotely (from the central office or street cabinet) or reversely (from the end-user’s home, over the telephone line). In the latter case, special care needs to be taken to ensure that the power draw is spread fairly between the active users, and that the system remains operational, even if all-but-one of the users have powered down.
Get to Fast, Faster
G.fast will play a critical role in many operators’ ultra-broadband network strategies, delivering more bandwidth to more people, sooner. For the operator, G.fast will result in a faster time-to-market and improved returns on investment. For policy-makers, it means that broadband targets can be met more easily, helping bridge the digital divide in terms of higher access speeds. For end-users, G.fast can help end-users enjoy new services or socio-economic benefits, sooner. In the end, G.fast will change the lives of millions of people around the world for the better — and that’s what counts.