Virgin Media O2 UK Start Selling XGS-PON Powered FTTP Broadband

XGS Is On says:

Same as regular products.

Anon says:

Thank you both.

Roger_Gooner says:

@Flame Henry: There is so much wrong with your post. For starters, VM’s HFC network is FTTN not FTTC, there’s a difference. And quite why you think that “VM02 have had the ability to go PON from the very beginning” when VM didn’t introduce fibre until late 2015. As for RFoG: this was entirely sensible as an interim solution to keep STBs and hubs in use, and all will be changed with the introduction of XGS-PON and going all-IP.

XGS Is On says:

‘The costs of D3.1 were huge’

No they weren’t. In most areas the basic implementation for Gig 1 required no work outside of hub sites and headends.

The discussion to go full fibre has been an ongoing one since the early 2000s. Openreach’s behaviour has not been a major consideration: there is still HFC cable in nations where the incumbent skipped FTTC/VDSL entirely.

The major needle movers are PIA and DoCSIS 4.0’s much higher cost than 3.1. To use 4.0 properly requires nearly all the network to have new amplifiers, new optical nodes and even new tap banks. Some of these costs would’ve been consumed with 3.1 had VM deployed high split but they did not.

FTTP did not make sense in 2012. 3.1 was not extortionately expensive, most of the network was upgraded over a long period and even now 3.1 is a bare bones implementation that didn’t cost much.

I’m sure some commentators disagree, especially those that don’t think hybrid networks should’ve been a thing at all, but the cost of 4.0 compared with 3.0 and 3.1 is a different matter entirely. Just to implement basic 4.0 with no increase in upload you’re close to the cost of FTTLA as a cable company either needs to move everyone to a 3.1 modem or replace amplifiers and optical nodes.

When that’s done in order to use 4.0 anything like properly nearly all the network needs new amplifiers, optical nodes, tap banks, splitters and isolators within the home to move to high split and 1.8 GHz.

XGS Is On says:

So much wrong in one post however it should be pretty epic to see what Roger comes back with.

If there were a subscribe button on this thread I’d be smashing it.

Flame Henry says:

@XGS_is_On – Care to highlight the bits that are wrong?

All of the responses to my comment further highlight the issue in that DOCSIS and HFC is presented as the only path open to them.

When people discuss HFC, they nearly always repeat the same arguments from the US and Germany where Cable companies are HFC only. VM was never HFC only. It had a huge network of fibre-connected cabinets full of equipment not too disimilar to the XGS-PON ISAMs they are installing today. Unlike most cable companies, they weren’t starting on a journey to PON from scratch. They already had systems in place to operate these kinds of networks but they neglected and ignored them.

Instead, they threw more and more money and effort upgrading a network that was rapidly becomming legacy. It’s taken 10 years and many Billions for them to realise they were wrong (and we, the public, have been denied true FTTH as a result). And all for what? So they can completely replace it with a network of ISAMs full of PON cards with not a trace of DOCSIS left. What was preventing that from happening 10 years ago when all of Scandinavia, Japan, Korea, etc. were already sitting on 1Gbps connections.

@Andrew G is right, they took an incredibly short-sighted bet that DOCSIS would outlive PON.

XGS Is On says:

‘@XGS_is_On – Care to highlight the bits that are wrong?’

In a rush so will miss stuff.

‘That MSAN was only ever used for voice services, but it was capable of so much more. Even in the early 90’s when they first deployed the MSANs, they were connected by Fibre (PDH/SDH), hence, they have had an “FTTC” network for decades.’

The MSANs were not capable of handling PON cards. They were PDH/SDH kit for aggregating and digitising phone calls. PON wasn’t a thing when most of them went in and they’ve neither the backplane or software to support it.

‘Did they swap out their POTS cards for PON cards? Of course not, they connected them to the cabinet nextdoor which contained a clunky HFC ‘fibre’ node ‘

The HFC ‘fibre’ nodes have always had fibre to them. HFC literally stands for ‘hybrid fibre coaxial’. The only new fibre going in was business as usual capacity upgrades. A very few areas, such as Leicester city, had no fibre to the coaxial network and were remedied in the early 2000s.

Fibre needed because of physics – distances to head ends and hub sites too long for metallic links, which is why the MSANs were there in the first place too.

‘90% of the bandwidth was required to deliver endless, crap TV channels (another terrible failure to acknowledge a superior technology (Video Streaming) was taking over).’/

The networks carried TV from the beginning. First analogue then digital. The CATV on manhole covers and some cabinets is short for ‘Community Antenna Television’. Internet tacked on later. Digital transition freed up so much capacity it wasn’t an issue bar isolated areas with restricted HFC networks until the 2020s.

‘So the lowly MSANs sat there doing pretty much nothing as telephone usage plummeted. Their Ethernet fibre backhauls still connected and unutilised’

The backhauls weren’t Ethernet. They are PDH/SDH as expected from a POTS system. Wanting rid of this legacy equipment with the support contracts and maintenance costs is the major driver for VM to shift people to voice.

‘With cable no longer the king it once was, VM02 did the unthinkable… started installing Telecoms equipment into their street cabinets and using it to deliver Ethernet services.’

That’s been a thing for years. Business parks, etc, have said equipment in their cabinets used to deliver Ethernet services.

‘VM admitting defeat and building the same network as OpenReach.’

They aren’t building the same network as Openreach. They’re using street-side OLTs throughout rather than exchange buildings, and moving the bulk of the IP network into cabinets too, to deliver a Converged Interconnect Network where Ethernet-whatever services can be attached. It’s a PON network, yes, however exactly the same would be the case with MSANs. Just different cabinets and more or less of them.

FWIW making alternative uses of the MSANs for FTTC was considered, I believe very small scale trialed, required extensive MSAN upgrade, and proved unviable. For each new MSAN that could’ve had those or PON cards installed there were hundreds that couldn’t.

Flame Henry says:

@XGS Is On – Thanks for the comprehensive reply.

I believe you are right on the details, but my wider point still stands.

VM could have relatively easily adopted PON technology many years ago, much easier than many other Cable companies. But they chose not to make any serious changes to their business or confront awkward truths until they were forced to.
The very fact they are now rolling it out suggests it was entirely possible all along, they just chose not to. Instead spinning up a multi-billion project to roll out new HFC kit and endless terrible CPE devices, instead of seriously investing in an all-IP platform, like many operators have had for many eyars.
And guess what, the new XGS network doesn’t co-exist with DOCSIS at all, which is exactly the situation they would have been in had they upgraded their MSANs to PON all those years ago.

Should we have expected anything different from a company with a leadership team who can’t decide if they are a TV studio, a cable company or a device manufacturer and effectively owned by the US cable industry? Probably not. But that doesn’t excuse the fact they have sat on their hands for many years and conciously chose to continue investing in DOCSIS instead of embracing a superior technology and making necessary business changes.

As for not being the same as Openreach, at least VM are deploying XGS which is more than can be said for OR. At least when everyone is running the same technology, we can look forward to a future where operators aren’t constantly trying to out-do each other with pointless ‘faster-than’ speed claims. Just release an 8Gb product and be done with it.

XGS Is On says:

‘At least when everyone is running the same technology, we can look forward to a future where operators aren’t constantly trying to out-do each other with pointless ‘faster-than’ speed claims. Just release an 8Gb product and be done with it.’

No chance. Ofcom would be all over them, people would be complaining about their service not hitting 8 all the time, and the CPE would be really expensive.

‘And guess what, the new XGS network doesn’t co-exist with DOCSIS at all, which is exactly the situation they would have been in had they upgraded their MSANs to PON all those years ago.’

XGS co-exists fine with DoCSIS over RFoG. RFoG downstream is 1310 or 1610 nm, upstream is 1550 nm. XGSPON uses 1577 nm downstream, 1270 nm up.

Alex A says:

VM doesn’t use MSAN cabinets, it uses telephone multiplexing with either copper or optical links to the exchange where the voice equipment is. The POTS ports on routers send it as an RF signal (last time I checked, might have changed by now for XGS PON) to the same voice equipment.

Even if they did im not sure what the FTTC point is, VM could already do those speeds over HFC. Its doesn’t have to share its POTS lines. BT wouldn’t get subsidy where VM already provide higher speeds, it would have been a commercial deployment.

Flame Henry says:

I tried to avoid too much detail. MSAN stands for Multi-Service Access Node. So VM do deploy MSANs but don’t use them to provide data services, like OpenReach do for VDSL. VM just use POTS cards and yes, the backhaul may be STM in some cases and not Ethernet. But the whole point of those type of chassis is that you can change the function of it by adding different line cards.
ASFAIK VM are using Nokia ISAM OLTs for XGS. Well guess who else use the ISAM platform, OpenReach, and they use it to deliver VDSL in Cabinets (and GPON from Exchanges).
So VM could have easily upgraded their Telco network to unlock 10G PON services. Instead they retired it, spent a whole bunch of money on DOCSIS upgrades, and then realised they had made a mistake and are now spending a whole bunch of money rebuilding a cabinet based “Telco” network, which is exactly what I’m saying they should have done in the first place 10 years ago when FTTH first became viable.

The fact that they are using PIA now helps them lower the cost, but wasn’t a reason not to do it before. They were doing the fibre dog anyway for RFOG, why not have made it XGS.

As for the co-existence, if they are intending to run both XGS and RFoG (not my understanding) to the same home, then this also would have been possible 10 years ago and thus, even more perplexing why it’s taken them so long.

XGS Is On says:

‘So VM could have easily upgraded their Telco network to unlock 10G PON services. Instead they retired it, spent a whole bunch of money on DOCSIS upgrades, and then realised they had made a mistake and are now spending a whole bunch of money rebuilding a cabinet based “Telco” network, which is exactly what I’m saying they should have done in the first place 10 years ago when FTTH first became viable.’

VM couldn’t have upgraded the MSANs. They looked into it years ago. They were considering it as a way to offload very heavy users from the shared HFC network and abandoned the idea.

They can’t just cut people off from phone service for hours while they replace MSANs, repatch the phone lines into more dense line cards to free up slots, taking care to do flawlessly, requiring rearrangement of the copper and cable management, and install PON cards.

I’m not sure where this idea comes from but it needs to die. Using the existing MSAN cabinets costs more due to the need for the above and is disruptive. Far easier to install a new cabinet every 3,000 homes.

If they could have and it made sense at some point they would have.

XGS Is On says:

‘As for the co-existence, if they are intending to run both XGS and RFoG (not my understanding) to the same home, then this also would have been possible 10 years ago and thus, even more perplexing why it’s taken them so long.’

It’ll be on the same fibre even if a single property can’t have both.

They were waiting for CIN switches to be good enough and for XGSPON ONUs/optics to become cheap enough. That plus improvements in install technology and lower fibre cabling and management costs and they were good to go.

They went RFoG because they could provision the same way as existing HFC, use the same CPE and the same CMTS.

XGS Is On says:

‘But we’ll happily swap every single optical node and do huge CMTS upgrades.’

As soon as they’ve been in the situation of having to swap nearly every single optical node they’ve gone for PON. DoCSIS 4.0 is the biggest upgrade by some way. CMTS upgrades are what they are – business as usual for cable companies. There are far fewer CMTS than there are MSANs or will be OLTs. The last round of upgrades collapsed 30+ CMTS in one site into 3 or 4. Tens of thousands of premises passed by a single CMTS now compared with about 3000 for an OLT and 500 for an MSAN area.

‘“But… fibre management” Not a problem. They have a solution for RFOG, why would it be different for PON.’

My point was it precludes use of an MSAN. I’m not sure why you’re so into those but they are not and have never been a viable upgrade route for VM to FTTP. Even the much more modest route to VDSL didn’t work.

Much as I’d have loved them to go full fibre years ago it wasn’t necessary and didn’t make sense. There was enough mileage in the coax. We should all remember that the £100 per premises passed is just to get the fibre to cabinets. Connecting customers, most of whom will already have coax to them I might add, is another £200+.

Flame Henry says:

“MSAN. I’m not sure why you’re so into those but they are not and have never been a viable upgrade route for VM to FTTP.”

A truly bizzare statement seeing as they are literally deploying Nokia ISAM hardware for XGS-PON… which is exactly the same platform that OpenReach use for their MSANs. Maybe my terminology is slightly off but you get the point. An XGS OLT is basically the PON name for an MSAN, to the point where it’s the same hardware platform underneath, and a whole world of difference away from a CMTS.

HAGWE

Alex A says:

@XGS Is On nevermind, reading your later replies we both agree that their cabinets couldn’t support PON linecards anyway but i don’t consider remote concentrators MSANs.

@Flame Henry the Nokia ISAMs they are using might support phone lines but the previous phone equipment are unlikely to support any PON line cards (like the Lucent Stinger I mentioned above).

VMs old cabinets also depended on the operator who originally built it before the mass of mergers. Part of the XGS PON rollout is further standardising the physical infrastructure (taps, the actual cabinet and whatnot).

Also the new chassis aren’t that expensive, it’s the line cards which they’d need to buy anyway is they could reuse the existing chassis…

XGS Is On says:

Ah yes indeed. Those I forgot about: they were in SDH mode transporting telephony when I had a loose involvement. I stand corrected on the IP capabilities of the newer kit, I only ever saw SDH coming out of it.

So, hands up, I was wrong.

PON wasn’t available on that kit at that time. Active Ethernet only. VM weren’t going to string a fibre to every home with DoCSIS 3.0 going strong.

I’m also far from sure they replaced every MSAN or even close to. 8500 is maybe a third of the fleet?

I guess you’ve had some involvement with MSANs at some point however I’m still stuck on where you think the business case was to abandon perfectly acceptable technology and run with inefficient options for GPON.

You wanted VM to replace the coax when it was delivering double what the competitive was without breaking a sweat. That makes no sense.

It also wouldn’t have relieved them from the expense of DoCSIS. It’s a 6 year project to get it available everywhere then a decade to get every customer installed. During that time it’d still need upgrade and VM would’ve had to upgrade CMTS to find space for more kit in the network.

Now, the kit’s backplanes are woefully underpowered for XGSPON so they’d all need replacing, too.

The timing was right and the method of deployment is right. Both having been decided by folks smarter than us.

I’ll actually ask one of the people who’d have made the decision why they didn’t use the MSANs when the GPON cards were released and get the full story. I’m quite sure it’s financial and business case.

XGS Is On says:

Flame Henry: One minor issue in your plan for VM – the MileGate 2300 and 2310, the two that support telephony, don’t appear to support passive optical networks.

That potentially explains why VM only tried VDSL on them.

I took a few earlier to have a look at a 2020 document on them and there is nothing in there about support for PON. Need the ones that are all-IP and do not have support for POTS.

Every day is a school day.

greggles says:

I am not VM’s biggest fan but I dont think D 3.1 was a waste of time.

VM have successfully rolled out gigabit speeds on it to their cable footprint, remember most consumers only care about download, and even the upload speeds arent far of Openreach FTTP for circa 94% of VM’s cable footprint right now.

What was a waste of time by comparison was g.fast, but luckily Openreach realised that mistake quickly.

Meanwhile circa 50% of Openreach’s footprint is stuck on VDSL or ADSL with a fair amount of that footprint “no existing plans” for Openreach FTTP.

In my city I expect OR to lose a ton of customers as CF are about to activate 2/3 of their remaining unactivated areas and of course VM have already rolled out gigabit services. Whilst OR have VDSL and ADSL services to compete.

XGS Is On says:

‘A lot depends on what the competition is doing, so no point wasting money to offer 1Gbps symmetrical when the biggest competitors are nowhere close to that.’

Offering 1 Gbit/s upload will cost VMO2 incrementally absolutely nothing over XGSPON. They are purely offering what they are to avoid the folks on HFC and RFoG getting butthurt and it creating marketing issues when it highlights that there are inferior DoCSIS areas and superior PON areas.