nbn tries to shift the conversation to future copper upgrades

In the best possible world, all terrestrial internet connections would use fibre-to-the-premises (FTTP). Everything I've learned over years writing about broadband suggests that a fibre optic network has a longer working life and will scale to greater bandwidth than technologies that rely on existing or new twisted pair copper …

[mathew42]

Chicken & Egg Problem

> But it [NBNCo] doesn't feel that FTTP is the way to meet that future demand. Indeed it doesn't believe there's a time at which gigabit-per-second home services will be necessary, but does believe that once such demands emerge they can be served with heirs to G.fast over copper.

Labor had two choices when establishing NBNCo:

- Expensive connection charges (AVC) for speed tiers and cheap data (CVC)

- Cheap connection charges (AVC) and expensive data (CVC)

Instead they chose the middle ground of speed tiers (AVC) and close to expensive data (CVC).

Labor predicted the result of this decision would be 50% connecting at 12Mbps on fibre. Currently 33% are connected at 12Mbps and 46% at 25Mbps. The 46% on 25Mbps is more than double Labor's prediction and this is most likely due to Telstra not offering a 12Mbps plan.

The result of this is that innovation has stalled because 79% are choosing the slow plans that don't offer access to the 100Mbps required for the new applications promoted by Labor in the NBNCo Corporate Plans. 1Gbps plans are available from NBNCo but RSPs are not selling them because a viable price point is simply not possible. Worse still congestion exists at the interconnect between NBNCo & RSP networks because of the high CVC prices. Labor succeeded in building a 1Gbps network on which nobody can connect at that speed because it is too expensive and a network that has extensive congestion issues because of high CVC prices.

If Labor had chosen cheap connection charges and no speed tiers, then connection speeds would be less predictable because of congestion issues especially during peak time but burst speeds could reach 1Gbps. RSPs could have innovated by offering higher quality services (e.g. plans that prioritised packets) or traffic shaping based on usage (e.g. old Internode FlatRate plan). End-users would have the option of either shaping their behaviour by using the internet in off-peak times or paying extra for higher priority.

[melts]

Re: Chicken & Egg Problem

I'm going to be upfront, initially your view just annoyed me, as usually it appears to be coupled with an argument against fibre to the premise as a solution

on pondering though, I thought well if you're argument is simply against the speed tiers, then I can understand that. problem is I think its better to have than not have, connection charges being factored in the same - lets assume that the price of CVC wouldn't change in a single tier mode and to offer a gigabit connection the AVC was at the 100mbit price if not more, and it'd be the only AVC available.

I like tiered plans, as when it works you get what you pay for. and I don't see any reason to deny users choice.

What you seem to imply though is a FTTP user at 12mbit is stifling innovation. What I see is there is nothing on offer for the user to upgrade. The early adopters are already on 100/40 and it isn't like that group are a single digit percentage. If you have an innovative product they'll be the first to use it and talk about it anyway. (I can't verify your figures as I can't find them, and as I can't find any I won't quote any. Feel free to source me though.)

Why I like tiers is it helps plan for congestion and lets those who only need a basic connection have it. Sure I'd of got rid of the 12mbit plan as its not IEEE defined broadband but otherwise I'd fine with it.

Letting ISP's have a full speed AVC and them manage like you propose hasn't worked well for ADSL and I don't see it working well for faster links. People believe these companies when they say they offer unlimited whatever, and don't understand contention or QoS. Making it one big pipe and asking ISP's to manage it when they instead can just state the peaks as the connection seems like it'd drivethe cost up for a poorer experience. Especially as the current FTTP GPON won't do a full 32 active 100 mbit connections without contention at the delivery level. (but hey at least it can be upgraded...)

I don't see why you obsess over the current connection speed at all really. I don't know a single ISP who won't let a customer go to a higher plan mid contract. if they have FTTP they know they can do it too. Maybe if if instead of saying massive percentage of FTTP customers are connecting at slow speed x, we say all FTTP customers are gigabit capable. Does that whet your thirst for speeds?

If you're saying it should be single speed for uniform pricing you're really saying the price should go up for those who don't need the same speeds, or do I misunderstand your thoughts on removing tiering?

[mathew42]

Re: Chicken & Egg Problem

> lets assume that the price of CVC wouldn't change in a single tier mode and to offer a gigabit connection the AVC was at the 100mbit price if not more, and it'd be the only AVC available.

NBNCo require a fixed sum of money to build the network and maintain it. NBNCo delivering a return on investment is heavily dependant on ARPU rising to over $100/month, and the growth strategy is total CVC revenue increasing. If you read the NBNCo Corporate Plan, CVC pricing started $20Mbps/Month when the average data usage is 30GB/Month and over time will fall to $8Mbps/Month when the average data usage is 540GB/month. This means that while the price falls by 2.5 times, while the average data usage grows by 18 times = growth in revenue from CVC of 720% when accounting for price falls.

Currently AVC is sufficiently high that for low demand customers with a mobile it doesn't make financial sense. If AVC was reduced to a flat rate ($20/month) and data charges were increased (or even just left at the current pricing) then demand for connections would jump and people would download more simply because the faster speeds mean people download richer media. NBNCo might take a hit in the initial years, but the growth in data consumption would rapidly see this eclipsed.

> The early adopters are already on 100/40 and it isn't like that group are a single digit percentage.

At the current drop in 100Mbps users (down 3% in 12 months) the percentage may well fall below single digits based on the current trend. The source for the information is easy to find in the media release section of the NBNCo website in the "nbn Half Year Financial Results Presentation 2016" presentation.

http://www.nbnco.com.au/corporate-information/media-centre/media-releases/Strong-result-continues-nbns-momentum-to-full-year-targets.html

> Letting ISP's have a full speed AVC and them manage like you propose hasn't worked well for ADSL

It has worked brilliantly compared with Telstra's model of charging for speed tiers and also having quotas. As with everything you do need to pay a little more for quality.

> Especially as the current FTTP GPON won't do a full 32 active 100 mbit connections without contention at the delivery level. (but hey at least it can be upgraded...)

GPON2.5 will support ~78Mbps with 32 users. If so much data was being downloaded, then NBNCo would be rolling in cash and the upgrade to GPON10 would be trivial to justify to the bean counters.

> Maybe if if instead of saying massive percentage of FTTP customers are connecting at slow speed x, we say all FTTP customers are gigabit capable. Does that whet your thirst for speeds?

I can tell you my car is capable of 350km/hr, but that is meaningless if the speed limit is 25km/hr. Continuing the car analogy, Labor's financial model is to build a freeway with toll charges based on the speed limit on the lane you select. The price for the 100x faster lane is only 4x the price but only a handful of drivers can justify the prices.

Labor's prediction is that in 2026 1% of customers will be connected at 1Gbps. Currently I'm not aware of RSPs offering a plan faster than 100Mbps, because it simply isn't financially viable.

> If you're saying it should be single speed for uniform pricing you're really saying the price should go up for those who don't need the same speeds, or do I misunderstand your thoughts on removing tiering?

I'm suggesting that AVC prices should be cut to a single speed tier and CVC prices should fall slightly slower than predicted in the NBNCo Corporate Plan.

If 79% of customers are connecting at 25Mbps or slower then the choice of technology doesn't make a perceptible difference. The small minority who want fast speeds have the option of moving to a FTTP area or fibre on demand.

[Jasonk]

Re: Chicken & Egg Problem

"GPON2.5 will support ~78Mbps with 32 users. If so much data was being downloaded, then NBNCo would be rolling in cash and the upgrade to GPON10 would be trivial to justify to the bean counters. "

But then it's also higher than that as only 26 get connected with 8 spare for other poodle upgrades.

Confusing speeds with data. The fact is someone would get what they want 4x faster than the current design of only up to 25Mbps.

"Continuing the car analogy"

So coalition is building for $56B a freeway that it's only required to do 25Km/hr to 100km/hr Some will be able to go faster but not some or most without have to pay costly upgrades ( one quote for 300m FOD was upto 10k almost 3 times the FTTP cost). According to the SR for $64B for s freeway that supports 100km/hr if people require to drive at that speed no need to pay for costly upgrades as everyone has a choice if they want to drive at that speed or not.

Skymesh is offering a 100/100 on FTTP which is really a 250/100 connection. Can HFC or FTTN do 25/25.

"I'm suggesting that AVC prices should be cut to a single speed tier and CVC prices should fall slightly slower than predicted in the NBNCo Corporate Plan."

So what you say s to adopt the current ADSL priceing where people getting 1Mbps paying the same as people getting 15Mbps.

[mathew42]

Re: Chicken & Egg Problem

In the article "Adelaide as a smart city" (http://www.abc.net.au/news/2016-03-19/adelaide-smart-city-privacy-concern/7259596), from which I quote:

> He said one area where integrated technology might improves lives was for an ageing population where people wanted to stay in their own homes for as long as possible.

> "Elderly people would be able to interact not only with the outside world with minimum support and supervision," he said of sensors and other technology that could be deployed.

For the interaction to work you need 100Mbps plus, but an elderly person on the pension is unlikely to have chosen a 100Mbps connection because of the extra cost. The impact of this is that they see the options demoed in house the video call quality is poor and so they don't use the system. If they had 100Mbps then they can self limit their video calls in the same way they self limit their electricity usage.

> if people require to drive at that speed no need to pay for costly upgrades as everyone has a choice if they want to drive at that speed or not.

Currently a small number 16% and falling are demanding a government service but don't care if others can afford that same service. My position is the fast speeds should be available to everyone and if not then people should be expected to pay for their fast internet. The cost of internet services (like other utilities) should be primarily based on usage, not connection fees.

> Skymesh is offering a 100/100 on FTTP which is really a 250/100 connection. Can HFC or FTTN do 25/25.

Except it isn't being sold as faster than 100Mbps.

> So what you say s to adopt the current ADSL priceing where people getting 1Mbps paying the same as people getting 15Mbps.

Your focus is on speed, whereas it should be on data. Both people are paying to access the same amount of data.

[Jasonk]

Re: Chicken & Egg Problem

"Currently a small number 16%"

So if the % stays that would be around 1.6m users not really that is it.

"My position is the fast speeds should be available to everyone and if not then people should be expected to pay for their fast internet."

But fast isn't available to everyone. Why should some have to pay while others don't? HFC getting a free upgrade to 3.1 but users are not paying for that upgrade.

"The cost of internet services (like other utilities) should be primarily based on usage, not connection fees.".

Only works if you are supplying the same level of service to everyone. You get a water supply that you can only fill a cup of water an hour while you neighbor could run a washing mechine if they decided but only require to fill a cup of water which can be done in 15min instead of the hour for you. so your paying the same price with a different level of service. If you want to run a washing you have a very costly upgrade for that choice while your neighbor doesn't.

Lol is solid as one but it's still a plan higher than 100Mbps. But FTTN and HFC being unable to provide that same level of service.

"Your focus is on speed, whereas it should be on data. Both people are paying to access the same amount of data."

Hang on 78% using 25Mbps or less. So what you are claiming is we should go back to 56k because speed doesn't matter that way they can download as much data they want no need to $56B for an upto 25Mbps (1Mbps faster than ADSL2) when only $8B more for a 100Mbps service. Doesn't matter that it might take them a week which could be done in acouple of minutes.

[mathew42]

Re: Chicken & Egg Problem

> Only works if you are supplying the same level of service to everyone.

Guess what water pressure varies also. If you are close to a distribution point, then it is likely that 2-3 people can shower at the same time. If you are further away then only one person can shower and running the washing machine won't work.

> Hang on 78% using 25Mbps or less. So what you are claiming is we should go back to 56k

Firstly it is 79% choosing 25Mbps or slower.

Did you miss the example I gave of the elderly being able to stay in their homes longer with HD video conferencing? My point is we should set a minimum standard, and anything beyond that standard should be priced accordingly. 79% selecting 25Mbps or slower suggests that this is a reasonable standard, that the general public agree with.

Your argument is "I want 1Gbps internet, so we should build a FTTP network. The fact that 79% are connecting at the unacceptably slow 25Mbps speed is irrelevant." This argument has been rejected by the Australian public. If your argument had been "The minimum speeds required to achieve these benefits (e.g. eHealth, eLearning, etc.) is 100Mbps but 1Gbps is preferable" then FTTP would be the only option.

> $56B for an upto 25Mbps (1Mbps faster than ADSL2) when only $8B more for a 100Mbps service.

First it is a minimum of 25Mbps on FTTN and for most connections significantly faster. Your need to retreat to 56Kbps arguments shows up that we do need a minimum standard for speed. The question is should it be 25Mbps or a faster speed?

Second the $8B is tax payer funds, whereas fibre on demand is your own hard cash. Compare this with the health system where the public hospitals set a minimum standard and people can opt for private health care which offers more comfort and better service.

http://theconversation.com/what-are-better-public-or-private-hospitals-54338

> Doesn't matter that it might take them a week which could be done in a couple of minutes.

Actually it does matter because just about all the benefits touted by Labor as reasons for building the NBN were real-time, not volume of data downloaded. Video conferencing is a prime example of this.

[Jasonk]

Re: Chicken & Egg Problem

Yes I saw your example. How about real time monitoring of those that need it. What would that require. What would the next communication app require in 5 years time.

"First it is a minimum of 25Mbps on FTTN and for most connections significantly faster."

Lol I didn't know the new min standard is to supply a 25Mbps for 1 sec in a day. During the 18Mth transition period that requirement is only 12Mbps for 1 sec a day. Looks almost like our current ADSL up to 24Mbps service doesn't it. Worst part is if your paying for 100Mbps and you still hit that 25Mbps for 1 sec in a day your connection is fine.

"Second the $8B is tax payer funds, whereas fibre on demand is your own hard cash"

First labor cap investment was $30B the current cap investment is now $29B. Rest of the fund coming from the private sector. So we are only saving $1B not going FTTP from taxpayer investment.

Second it's a loan not a budget item. So what your saying is the hospitals just now make a return for the government to pay back any money the govenment pays them.

Third if we use the SR $64B price tag which only $8B more than the mess we have now. To upgrade FTTN later if we use current rollout figures it's $16B. If we use the FOD price of $10K + it's $45B more.

"Actually it does matter because just about all the benefits touted by Labor as reasons for building the NBN were real-time, not volume of data downloaded. Video conferencing is a prime example of this."

First you claim in another conversation that speed doesn't matter now it does which is it.

[mikeinnc]

Re: Chicken & Egg Problem

:Second the $8B is tax payer funds, whereas fibre on demand is your own hard cash."

Except we haven't seen a single fibre on demand connection. Not one. It's been touted; it's been talked about but it is as ethereal as the fairy at the bottom of your garden. The current FTTN network architecture cannot support it, and NBN Co really do not want to supply it. It's just another blatant lie from an ex-communications minister who nose must exceed the length of Pinocchio's by now!

[Simon Sharwood, Reg APAC Editor]

Re: Re: Chicken & Egg Problem

"Only eight billion ..."

That's the bit I choke on. That's eight billion on which interest has to be paid.

[Abel Adamski]

Re: Chicken & Egg Problem

With respect over a multi decadal infrastructure project that is a National Interest project of major import to the economy, that $8Bill will be dwarfed by operational and maintenance and upgrade costs of the dogs breakfast we are getting, especially with all the field mounted powered electronics as Climate Change is delivering increasingly weird weather, massive intense flooding and periods of intense heat.

Not to mention limiting the whole network, with massive amounts required for piecemeal upgrades.

Truly an engineering project designed by suburban tax accountants

[melts]

Re: Chicken & Egg Problem

I like what you had to point out but wanted to address a few things;

> For the interaction to work you need 100Mbps plus, but an elderly person on the pension is unlikely to have chosen a 100Mbps connection because of the extra cost. The impact of this is that they see the options demoed in house the video call quality is poor and so they don't use the system. If they had 100Mbps then they can self limit their video calls in the same way they self limit their electricity usage.

100Mbit is probably correct. and OK its four times the cost (if I recall your maths correctly). How much is it compared to a care worker? This is where you subsidise things, and the company offering the monitoring would be involved in this, as well as service provision really - this is the ultimate in managed services, isn't it. Arguing the price of the connection in a vacuum isn't reflective of the total costs, and looking after someone costs substantially more than a 100Mbit connection

> Currently a small number 16% and falling are demanding a government service but don't care if others can afford that same service. My position is the fast speeds should be available to everyone and if not then people should be expected to pay for their fast internet. The cost of internet services (like other utilities) should be primarily based on usage, not connection fees.

while I agree that everyone should have equal access, ISP's have proven it won't happen. You even stated you need to pay a bit more to get a better service. Whats the point of everyone has 100Mbit if its over saturated and doesn't provide when you want it, ie peak loads

If power and water were managed like bandwidth we'd be in trouble. Ultimately if that was fixed your view would be fine.

> > So what you say s to adopt the current ADSL priceing where people getting 1Mbps paying the same as people getting 15Mbps.

> Your focus is on speed, whereas it should be on data. Both people are paying to access the same amount of data.

problem being the person on 1Mbps can't consume the same amount of data. a lot of data services are throughput constrained. even if the connection was free and they only paid for data the 1Mbps person is going to be very unhappy as he can never stream anything past a radio channel, no netflix for him.

FTTP means you *can* get the same link speeds, and you can pay more for a better link (like you can bond ADSL connections, if you so desire, except, reliable and affordable). then you do battle for your congested bit of data.

removing the tiers would be fine if the ISP's could handle the demand, they can't at the moment, so asking people to either pay more (AVC would need to go up to be universal, surely) to get a poorer service, or those already paying more to have maybe a drop in AVC price, but get a much degraded service, seems a bit harsh.

at least I like the idea of fairness.

[melts]

Re: Chicken & Egg Problem

firstly, thanks for the figures, possibly my lack of keywords but I didn't find that via Google earlier.

> NBNCo require a fixed sum of money to build the network and maintain it. NBNCo delivering a return on investment is heavily dependant on ARPU rising to over $100/month, and the growth strategy is total CVC revenue increasing.

While I would argue about the financial model and how to fund it, it'd be off topic. Assuming we want to aim for that figure to meet their requirements, you're plan is reasonably well thought out, except it relies on ISP's to manage the contention. If it wasn't for that I'd agree that a lower fixed connection price would be better to drive demand.

> At the current drop in 100Mbps users (down 3% in 12 months) the percentage may well fall below single digits based on the current trend.

(down from 19% to 16%)

I honestly don't know why the 100mbit group are going down, but I can imagine its because ISP's everywhere are struggling to deliver it in peak periods, that is what I hear being in an area going FTTP. It makes me sad really, as people associate FTTP with bad service, unable to pinpoint their problems on their shitty ISP. I don't see that being fixed by abandoning tiers though, if anything I think like a congested DSLAM it'll just make the problem worse on peak

> It has worked brilliantly compared with Telstra's model of charging for speed tiers and also having quotas. As with everything you do need to pay a little more for quality.

it hasn't though, as the sales pitch never explains you need to pay more for quality. its not what the ISP's selling their wares are going to explain. I'm not sure what Telstra model you're referring to, but I am aware of ISP's over selling their capacity and never explaining contention. What is it I'm missing?

> GPON2.5 will support ~78Mbps with 32 users. If so much data was being downloaded, then NBNCo would be rolling in cash and the upgrade to GPON10 would be trivial to justify to the bean counters.

and I agree, I don't see that being over subscribed at that level. (and as someone indicated it wasn't even 32 due to future proofing etc) I was simply highlighting there is that bottleneck and we don't look like we'll be approaching it any time soon. Unlike FTTN.

> I can tell you my car is capable of 350km/hr, but that is meaningless if the speed limit is 25km/hr. Continuing the car analogy, Labor's financial model is to build a freeway with toll charges based on the speed limit on the lane you select. The price for the 100x faster lane is only 4x the price but only a handful of drivers can justify the prices.

I'm confused why its an issue, you pay to get more, seems ok to me. But to go on the car analogy, I used max speed to show what the infrastructure is capable of. It doesn't have any upto asterisks on the speed limits, unlike other technologies in play. Could you imagine a FTTN freeway where you can't actually provide the speeds the cars will travel at until they give it a go, and yet will charge them the same amount to use it? Wouldn't work.

If it was a multi lane freeway and I paid more, but not too much more, to go faster, I would. If I was told I should pay almost as much as my car to get the top speed I'd have to re-evaluate my choices (oh but its not too much if you own a house I've heard you state, quoting cost of house vs FoD, it's a bit of a hollow argument. Like saying you only need to get a good job to get a house in the first place)

On top of that to use the new car you have to pay more, kind of sounds like a car really haha, but I'd rather tiers than a single risky option as a nation rollout.

Labor's prediction is that in 2026 1% of customers will be connected at 1Gbps. Currently I'm not aware of RSPs offering a plan faster than 100Mbps, because it simply isn't financially viable.

2026 is ten years away. ten years ago the iPhone didn't exist and mobile data was a luxury. Things change and as usage increases the plans will come online. Its not a large number so I'm not sure why you think it's not going to happen, unless you assume like I do ISP's would struggle to deliver 1Gbps to anyone. Even I think it'll be offered for enterprise in limited forms before the end of the decade.

> I'm suggesting that AVC prices should be cut to a single speed tier and CVC prices should fall slightly slower than predicted in the NBNCo Corporate Plan.

sounds great, but as I've said before I think you cripple user experience via the ISP then. I'd rather it be a little more expensive to get a faster link and force some form of speed controls on the network than let ISP's do it.

> If 79% of customers are connecting at 25Mbps or slower then the choice of technology doesn't make a perceptible difference. The small minority who want fast speeds have the option of moving to a FTTP area or fibre on demand.

My argument is that 79% can pay at any time to go to faster, put them on FTTN and they can't. Even if they are rolling in cash they need to go through the protracted process of getting FoD installed, and I've seen it being quoted far more than the $5000 figure stated my Turnbull.

You might believe its fine to move to a FTTP area, or pony up 20k to get FoD, but as one of the 30% of Australian's renting, moving isn't fun and I can't magic up 20k for a rental property improvement. Rentals don't get these kinds of improvements. It's a new digital divide and I don't see why it has to be perpetuated.

If getting rid of speed tiers would see the back of this larger divide I'd get on board and let the ISP's mangle away with contention. I know it won't though, as those upset by it at the moment are the small minority.

I think that is the bigger problem and arguing that current use of FTTP justifies rolling out FTTN is not good long term thinking. I don't think rejigging the CVC and removing AVC choice will overly affect FTTP uptake - as much as I'd like to see it increase, it'll just make it look like we have faster connections, while giving us more contention.

[Abel Adamski]

Re: Chicken & Egg Problem

Matthew, still on the same rant after 4 years , just wanting something for free - a 1Gb service at the cost of 12Mb so you and others like you can soak up every available bit of data at minimal cost to yourself and maximum inconvenience to everyone else.

Grow up

[John Tserkezis]

"Some suggest that nbn is deliberately misleading the public by insisting that FTTP cannot be delivered as cheaply as FTTN"

There is no misleading around here - there is no leading of any type around our place either - we have no idea when this "nbn" thing will come, if at all.

There is a distinct lack of any chatter at all. The so called crunch times of 2017 and 2018 are utter bullshit.

[GrumpyOldBloke]

2017 / 2018 might be when the chickens come home to roost. nbn co are predicting that they will have massive deployments of the MTM nbn by then and they get to see if income covers costs or if the high CVC charges and operations costs, the collapse in mining investment, the closure of car manufacturing plants and parts suppliers, the drop in Oz R&D, reduced workforce participation as the boomers start to retire, the increasingly litigious environment around copyright in Australia and h265 mean that people find the lowest tier plans are good enough or simply migrate to wireless / tpg / transact. 2017 / 2018 might be when we get to see if the nbn has survived the LNP's two+ years of mismanagement or if the delay was terminal and the only path forward is to sell it in pieces at a loss thus ending a decade of stagnation in Oz (non wireless) telecommunications investment.

[Medixstiff]

But what about powering the nodes.

Not once have I seen an estimate on the power costs for all those nice FTTN nodes they are installing, that will add to the eventual cost of the FTTN solution.

Power which would not be needed with FTTP.

[FrancisYoung]

Cost parity of FTTN and FTTP

The question to ask Mr Morrow, in order to assess whether FTTN or FTTP is "cheaper", is how much average revenue per user is being received from FTTN areas.

If it is less than the $43 ARPU coming in now from FTTP areas, then how quickly will FTTP be at cost-parity with FTTN, especially as data demand keeps growing?

Anecdotally, the break-even point occurs before the rollout is completed, and therefore urban areas should be done with FTTP wherever physically possible.

[Trixr]

Please grammar

"So yesterday, nbn drove I and several other journalists around Brisbane..."

Dude, it doesn't matter how many objects there are in the sentence - they ALL take the objective case.

"This phrase drove me spare..."

"nbn drove ME around Brisbane..."

"nbn drove me and several other journalists around Brisbane..."