G.fast and furious in 2018

Red car

Just when you thought the copper network was finally going to be retired and put out to pasture, engineers continue to find more ways to extend the life of this aged infrastructure.

Vectoring – perhaps one of the greatest technology enhancements ever developed for the copper network – has removed many of the issues and concerns related to using higher frequencies on copper pairs. This noise-cancellation technology allows operators to offer higher speeds in both the downstream and upstream.

G.fast further leverages vectoring technology along with time division duplexing (TDD) and higher frequencies to help operators get to that magical broadband benchmark of 1Gbit/s.

For many broadband operators the ability to offer Gigabit broadband services is the golden ticket to customer retention and higher revenues. In our annual Ultra-Broadband Global Operator Survey – the number one and overwhelming driver for the deployment of these new technologies is the ability to offer 1Gbit/s.

Key developments driving increased interest

In its initial specification, G.fast was optimized for very short loop lengths – with the fastest speeds (1Gbit/s aggregate) achieved at around 50 meters. However, since its initial approval in 2014, three follow-on amendments have spurred increased momentum and operator interest as new features and capabilities have extended both the rate and reach for G.fast.

Key developments that are creating operator enthusiasm for G.fast include the following:

212MHz profile

The 212MHz profile (double the spectrum of the original specification) allows for faster speeds up to 2Gbit/s (aggregate) – allowing for 1Gbit/s symmetrical.

Coaxial support

Amendment 3, Annex X defines a “c” (coaxial) profile – applicable to both the 106MHz and 212MHz profiles. This annex specifies operation of G.fast transceivers in a crosstalk-free environment – meaning vectoring is not used and synchronization between lines is not needed.

Dynamic timeslot assignment (DTA)

DTA is about adapting the DS/US allocation in a dynamic way based on the actual traffic requirements to improve end user experience (QoE). DTA dynamically allocates the aggregate capacity to the direction that best serves the instantaneous needs of the user’s applications.

Two variants of DTA are supported – independent (iDTA) and coordinated (cDTA):

  • iDTA: When there is no cross-talk between the lines (e.g., coax), each line can adapt its DS/US allocation independently of the other lines
  • cDTA: When there is cross-talk between the lines (e.g., TP), all the lines have to use exactly the same frame structure

cDTA, like the earlier iDTA feature, improves G.fast upstream performance by four to five times by dynamically balancing upstream and downstream capacity to match residential traffic patterns in real time. cDTA also expands the applicability of this feature to existing phone wiring, thus covering nearly all residential and commercial premises.

G.fast interoperability and certification

Interoperability and certification are a key hallmark of broadband services. As such The Broadband Forum in partnership with the University of New Hampshire’s Interoperability Lab (UNH-IOL) are moving quickly to certify G.fast products and accelerate the availability of interoperable solutions and expedite timely integration with service providers’ networks.

To date, the number of certified products has reached 24, including five different vendors certified for DPUs and seven for CPE.

What’s on the horizon?

There appears to be more bandwidth that can be squeezed out of the copper plant as a new recommendation – Multi-Gigabit Fast Access to Subscriber Terminals (G.mgfast) – is currently in the proposal stage with the ITU-T Study Group 15. G.mgfast could theoretically offer aggregate data speeds of 10Gbit/s per cable at 848MHz or 5Gbit/s at 424MHz.

The anticipated use case for this new approach is predominantly focused towards backhaul applications for small cells, Wi-Fi infrastructure, and other 106 or 212MHz G.fast DPUs.

An unexpected opportunity?

Virtually all G.fast deployment use cases have focused on multi-dwelling unit (MDU) deployments. In fact AT&T – who has been particularly bullish on its fiber-to-the-home deployments, announced in August 2017, that is was deploying G.fast within MDUs in 22 metro areas. While that announcement is not necessarily groundbreaking – the fact that each of these metro areas is located outside of AT&T’s traditional 21-state home internet service-area is.

For these MDUs, AT&T is deploying G.fast over existing coaxial cables, offering speeds up to 500Mbit/s/100Mbit/s, with future speed increases on the horizon. AT&T is pairing this G.fast broadband services with its DIRECTV service.

For its G.fast deployment, AT&T is using the 106Mhz profile over coax, and using GPON as its backhaul (with XGS-PON planned for the future) for its 16-port G.fast DPUs. DPUs are primarily deployed in an indoor equipment room, but also in an outside cabinet. Currently, local power with battery back-up used to power DPUs; however, RPF (reverse power feeding from CPE) is expected to be introduced in 2018.

AT&T is also expressing strong interest in the proposed 424Mhz profile over coax in order to offer 2.5Gbit/s symmetric service for strategic deployments.

Bridging the gap

G.fast is meant to “bridge the gap” – enabling operators to offer FTTH-like speeds where fiber is not currently available or time to market is of high importance in a competitive environment.

As shown by AT&T’s strategic deployments; key features such as support for 212MHz and coaxial cable have proven to be attractive characteristics of G.fast that are allowing it to be a viable option for their broadband services but also opening up potential deployment opportunities for a variety of operators.

For many operators, the deployment of G.fast is proving to be the preferred choice for offering gigabit broadband services to MDUs, as it eliminates many of the issues found with FTTH deployment such as the variety of building types, access to the riser, centralized versus distributed closets, and overall construction costs.

The availability of Amendment 3 features has been a key development towards growing interest in G.fast from operators across the globe – enabling them to both position their networks for the future (deeper fiber deployment), while continuing to leverage the existing infrastructure – particularly the last few hundred meters – in a cost-effective and competitive manner.

As such, we expect 2018 will prove to be the key year in accelerating G.fast deployments.

Related articles