Openreach FTTdp and G.fast Test Hits 1.66Gbps Aggregate Broadband Speeds

A couple of months ago we became the first to detail Openreach’s trial of a combined FTTdp and G.fast broadband network that used hardware from NetComm Wireless. Today NetComm has confirmed that their test achieved a peak aggregate broadband speed of 1.66Gbps.

Full details of the test setup can be found in our original article from June 2017 (here), although suffice to say that the Fibre-to-the-distribution-point (FTTdp) based network made use of several key enhancements to G.fast technology in order to boost its performance (e.g. it can use up to 212MHz of spectrum instead of the 106MHz of initial deployments and the G.fast nodes can be reverse powered by home owners).

At the time Openreach told us that they had achieved a top downstream speed of up to 1Gbps and uploads of up to 300Mbps (c.1.3Gbps aggregate), which was apparently achieved at a distance of around 40 metres (copper run from the G.fast distribution point and into your home). By comparison the original G.fast Proof of Concept trial from a few years ago was only able to deliver an aggregate of around 600-700Mbps at a similar distance.

Fast forward two months to today and NetComm Wireless has now made an official announcement about the trial, which claims that the peak speed actually went even higher and hit a 1.66Gbps aggregate at the same distance of 40 metres. In theory this approach, using the latest amendments to the G.fast standard, could thus deliver an aggregate speed of around 1Gbps when installed up to 150 metres from the premise.

Timo Brouwer, COO of NetComm Wireless, said:

“NetComm Wireless began developing FTTdp DPUs when it became clear that an alternative to fibre to the premises was needed. Openreach is always looking forward, and we are pleased to have demonstrated the latest generation of Gfast FTTdpcapabilities that will support the broadband demands of the future.”

Peter Bell, Openreach’s Director of Network Solutions, said:

“Our aim is to deliver high-speeds to as many people, as quickly as possible. We’ve outlined plans to provide ultrafast speeds to 12 million premises by the end of 2020 and we want to go further. We’re working with NetComm Wireless to understand how we can deploy this technology closer to the premises to reach even more customers.”

All of this is possible because in an FTTdp deployment you can build smaller G.fast nodes (dp) much closer to homes (e.g. on top of telegraph poles), which compares with the current approach of sticking the kit inside an extension pod that’s then fixed to an existing PCP street cabinet (i.e. many homes exist further away than ideal and the advantages of G.fast are thus significantly reduced).

The downside is that FTTdp based G.fast is quite a bit more expensive to deploy, albeit not as expensive or slow to deploy as a “full fibre” (FTTP/H) roll-out. This could make it an attractive option, albeit one that would be quite complex due to the need for lots of small nodes (at present each one can only handle 4 ports). We note that the G.fast nodes also include support for VDSL2, which means that existing FTTC lines might also benefit.

We’ve long heard talk of FTTdp making a return to help cater for areas that exist outside of the current PCP cabinet approach. In fact G.fast was always envisaged as being an FTTdp approach but once Openreach found that it was considerably cheaper, while still being quite fast (service speed), to deploy it from cabinets then it made sense for them to alter their plan.

However the future of G.fast is currently in some flux as Openreach consult on the possibility of deploying significantly more 1Gbps capable FTTP/H than originally planned. Evidence on the ground (see Monday’s news) shows that G.fast is still very much being deployed in the UK but it remains to be seen if the on-going consultation changes the operator’s future approach.

1. MikeW

   August 16, 2017 at 10:35 am

   Progress in research might mean that BT will wait for G.mgfast for the shorter range FTTdp solution. Aggregate rates of 4-8Gbps using 424MHz and 848MHz, over 35-75m distances.

   The current generation of 212MHz looks to still be of benefit to the BT’s existing plan with 300m ranges.
   • Mark Jackson

     August 16, 2017 at 2:00 pm

     Perhaps a bit too early to be talking about G.mgfast. I only saw the name mentioned for the first time by BT and ADTRAN a few months ago. Still seems to be in the very early research phase.
   • Matt

     August 16, 2017 at 2:40 pm

     Here was me thinking next thing after G.Fast was XG-fast lol clearly not paying enough attention.
   • MikeW

     August 16, 2017 at 3:34 pm

     Far too early, I agree.

     I think there’s plenty of time to pass before BT settles on its deployment strategy for deeper G.Fast infill, so time for research to mature too.

     I think XG-fast is what one manufacturer labelled their research (Nokia/Alcatel-Lucent), whereas G.mgfast seems to be the label attached to the ITU standardisation group. mg = “multi-gigabit”.
2. adslmax Real

   August 16, 2017 at 11:26 am

   FTTdp is bad move because it will not benefits all of us for any cabinet with more than 150m away. Totally worthless.
   • TheManStan

     August 16, 2017 at 11:47 am

     GFast FTTdp is not to cabinet, but to a node close to the properties being served…
   • Steve Jones

     August 16, 2017 at 11:51 am

     I’m not quite sure what part of fibre-to-the-distribution-point passed you by, but this article is not about g.fast from the PCP. It’s from nodes that might (for example) be mounted on a local telegraph pole.
   • Lee

     August 16, 2017 at 1:06 pm

     Think before you write max.
   • Ultraspeedy

     August 16, 2017 at 3:06 pm

     I thought it was to a new POD cabinet also rather than a pole.
   • TheManStan

     August 16, 2017 at 3:48 pm

     That is the commercial roll-out planned for later this year.

     OR have been trialling other GFast methodologies.

     I would hope this is the next stepping stone after their pod based roll-out and/or which would then be finally followed by fibre terminating in the customers property.
   • Ultraspeedy

     August 17, 2017 at 1:35 pm

     So based on a previous GFast story and the quote of “BT currently intends to make its GEA-NGA2 G.fast (ITU G.9700/9701) service available to 10 million premises by 2020, with “most of the UK” likely to be done by 2025 (we predict around 60% UK coverage).”

     This if they do it (after previously abandoning the pole route). Is likely to be 10 years away and another interim solution with more costs attached.
   • MikeW

     August 17, 2017 at 2:24 pm

     It has taken 9 years to get from the pilots of NGA1/FTTC to where we are now (ie still not finished with the rollout). Is it so unlikely that NGA2 would take a similar amount of time?

     It certainly has costs associated. However, because it also deploys to a more widespread area faster than an FTTP solution, it also has more income associated.

     I think Nesta had calculated (last year) that there was little difference in cost in deploying G.Fast now AND FTTP in 2023 vs just starting FTTP now, but there was a difference in when people got access to faster speeds. The earlier takeup offsets the costs.
   • Ultraspeedy

     August 17, 2017 at 5:47 pm

     What will happen to all the POD cabinets if they later go a mounted on the pole route?
   • MikeW

     August 18, 2017 at 1:21 am

     Best assume they stay. Even if they eventually only supply premises within 100m.
   • 8okeAKWiPe

     August 19, 2017 at 4:48 pm

     Please ban this idiot, adslmax. A nuisance wherever he turns up.
3. Matt

   August 16, 2017 at 1:55 pm

   FTTdp has always been considered good for everyone really it takes Fibre deeper into the network everyone who gets FTTdp would be able to benefit from G.Fast. Apart from cost to deploy it i can’t really see the downside.
4. MikeW

   August 17, 2017 at 2:28 pm

   I see that ZTE are talking about similar results when testing with NetCologne, reaching 1.6Gbps down and 200Mbps up.

   That too uses 212MHz, and is with Broadcom chips.

Comments RSS Feed