It’s now 11 years since the twinkle of the National Broadband Network was formed in the Broadband Advisory Group’s eyes. It’s like the current government is desperately trying to take us back to 2003, when Howard was Prime Minister, and Ziggy Switkowski was CEO of Telstra.
It was a great time for geeks: the HFC networks had just started to see decent saturation, and ADSL was almost at full saturation. We were moving from dialup so 10Mbps HFC and even 256Kbps ADSL seemed like they screamed. Not to mention the fewer dropouts.
We’re 11 years hence, and the plan has not changed much: the Coalition still want to use FTTN and still want to heavily involve Telstra in the process. Two decisions that have attracted the ire of many ICT professionals over recent years. Funnily enough, if the same plan had been committed to a decade ago, it would have been met with much enthusiasm from people like myself.
The key is, this is not a policy decision, this is a technical decision, and that’s where the Liberal party has tripped up over the last few years. Yes, the plan would seem smart, worthwhile, and a good investment, ten years ago. Why ten years ago? It’s quite simple: ten years ago copper technology was the most reliable for last mile transmissions without being prohibitively expensive.
In this day and age, most of the equipment used for VDSL2 is interchangeable with GPON gear, the main difference is the “card” that’s used, you just need a LOT more “exchange equipment” (this does include what’s in the FTTN “node”) for VDSL2.
Think of it like this: for a VDSL2 node to service 144 pairs you require 3 line cards just for the xDSL component (another 3 voice line cards, and another 3 “splitters” to combine VDSL2 & voice). For the same number of GPON customers you need 4.5 ports on a single line card. In the future, with TWDM-PON, the 1:32 split increases to 1:512.
This is all well and good, but why do I keep saying these FTTN/HFC polices are from the past?
I’m Sorry, Are You From The Past?
I keep mentioning on twitter that there are very few countries currently in the middle of FTTN or FTTC (Fibre to the Node or Fibre to the Cabinet, both essentially the same) rollouts. The two that spring to mind are the UK, and Italy. Yes, Belgian, Austrian, and Finnish telcos are trialling G.Fast via an FTTdp (Fibre to the distribution point, essentially “fibre to the pit”) set-up. You can’t really compare the two, and while I don’t think G.Fast is a wise investment at this stage, due to the regulatory minefield rolling out technologies that interfere with the FM-band, it’s definitely not in the same league as “maybe 25Mbps” we’re being offered by the Multi-Technology Mix (MTM) that has been proposed.
The key is, G.Fast (or Marvell’s G.now) is an end-of-line technology. It’s essentially the closest to FTTP you can get, without actually deploying FTTP. I’m not sure how the economics will work out, but a powered device in a pit is definitely going to cost more in the long run to a passive device on the side of a house. (Yes, I realise there are active parts of GPON inside the house, as are there with VDSL2)
The only technology that’s designed to deal with future fixed line bandwidth requirements is fibre. Not only from a user perspective of having reliable, flexible, broadband in the ultra-fast category, but from a commercial sense.
Investing in technologies that you know for a fact will start to struggle before you’ve completed the rollout is insane. How do we know this? We can see data use trends. Globally our data use doubles every 18 months, this has been observed for some times.
Even if we conservatively say that’s 7 times data use has doubled since FTTN was suggested back in 2003 by the Broadband Advisory Group, this means we’re using 128 times the data we were in 2003.
Let that sink in for a moment. By the time we the Coalition Broadband Network (CBN) is completed, we will be using 16 times the data we are TODAY. If there’s any delays, this starts to rise, just a delay of 3 years would mean we’re using 64 times the data we are today.
Which brings me to my next point: if you rejected FTTN 11 years ago, why are you deploying it to be ready in 6 years? Yes, that’s 17 years from concept to completion, providing there are absolutely no delays. With Telstra still not inking any lease/sale deals for the copper, expect this to drag on for some time. Remember, it took 18 months for Telstra to sign last time.
Back To The Future
What’s immediately apparent is that the current government prefers to consult consultants than experts when it comes to formulating policy, as there is a tsunami of FTTP deployment announcements from telcos the world over, while FTTN/C deployments are foundering.
BT appears to have bitten off more than they can chew, with rollout targets being revised down yet again, and even though OFCOM have already handed over £1.2 billion to them, they’ve come back with their hands out for another £250 million. This really doesn’t help prosecute the case for for FTTN. There’s even whispers of the FTTP component being increased dramatically in the UK.
Even traditionally wireless markets, African and poorer Asian countries, are seeing heavy investment in FTTP as it makes sense to service the emerging middle class. Although I do think some of the rationale is to avoid copper theft, which is a major problem in developing nations (it’s starting to be here too). Even with this ignored, it makes perfect sense to deploy low maintenance, high revenue fibre.
Why are they choosing technologies that conservatives claim will be “outdated soon”? You know, “5G blah blah blah”?
It’s pretty simple, the people making decisions actually consulted experts in the field. No expert in telecoms can truthfully state that FTTN is a wise investment in this day and age.
FTTP, even if ignoring the speed, is cheaper to run, consistent, and even increases house prices by up to 20% (as per recent research). These sound like positives, there must be negatives surely. Well, not really. Cost is about all there is, and even that’s fairly close. From all my analysis of corporate plans, reviews, and claims, I put the per premises difference in the $100-$200 range. Not much in it, which is why the CBN needed to go with HFC, to drop this cost and not invest.
What usually blows people’s mind is when you compare the current “in-development” technologies and their speeds, capabilities, etc.
G.fast or G.now (I’m not that familiar with the latter, it’s fairly “new”) are both ultra short run (<100m) DSL technologies, and while they do work on longer runs (up to 250m), you’ll never see full speed. Even at full speed you’re looking at 1Gbps, this requires vectoring at the very least and is dependent on the copper quality/gauge. One unit services 2 or 4 properties via 2 or 4 copper pairs.
Meanwhile TWDM-PON (Time & Wavelength Division Multiplex – Passive Optical Network) is the cream of the crop. Per wavelength you can squeeze out 10Gbps, the current test system is 8 wavelengths, so you get 80Gbps dedicated per customer. Each port services 512 customers with 80Gbps dedicated per customer. The upgrade to full speed TWDM-PON will require a little work, but moving to just 10Gbps would be the same as an upgrade to 10G-PON (10Gbps Passive Optical Network).
Where Back At the Beginning
The future is far from rosy for Australia, we’re seeing an unprecedented attack on technologically inclined people by our Communications Minister, lashing out at anyone who questions the validity of his CBN. This kind of ad hominem attack is a tactic used by people who don’t know what they are talking about.
I know the PM thinks Malcolm Turnbull invented the internet, but the evidence is that he doesn’t actually understand technology, nor does he understand why people are opposed to investing in 11 year old ideas.
There’s nothing new in the CBN, even the CEO of Telstra has been recycled from 2003.