Latest comments

Good observations, Bob.

Graham Shepherd's picture

Good observations, Bob.

In the latest Journal http://telsoc.org/ajtde/2014-12-v2-n4, I try and take an overall view of the latest policy and Gary McLaren compares the utility model you describe with the competitive model proposed by Vertigan. The government seems to be hovering between a utility approach and a comptitive approach. In the Business Spectator last week (http://www.businessspectator.com.au/article/2014/12/12/technology/year-nbn-stood-still) Mark Gregory noted that if the government delays selling off the HFC until NBN Co upgrades it then a well-cashed-up player like Telstra could walk right back in to a dominant position in broadband access.

I wish that rational argument would prevail but despite the mantra of rational economics I doubt that it raises even a lazy eyebrow. Daniel Kahneman in "Thinking fast and slow" demonstrated that markets and consumers are not rational (unless they really concentrate). I suspect that governments and politicians are the least rational of all and wide open to powerful forces, including half-cocked ideologies, political expediencies and modern, more sophisticated versions of the good old brown paper bag (Askins, I think they were called).

I remember Richard Alston laughing his head of at naive young me when I suggested that he simply had to explain to the Australlian people the "good reasons" for privatising Telecom Australia and he would get his legisltation through. Of course, it was the mountainous tax-payerfunded political pay-offs to Senator Harradine that actually ensured the result he wanted.

There will be a wide readership of the Journal articles. You might like to add your comments there.

Graham Shepherd

Gigabit connections: MyRepublic says NBN chief is wrong

Nigel Swinney's picture

Given the rediculous cost to get 100Mbits currently I doubt we could justify 10x that for gigabit but it would be good to have if they priced it right

(No subject)

Ian Campbell's picture

(No subject)

Rob Nicholls's picture

PDF file reissued

administrator's picture

Thanks, Bruce. I hope that you received the reissued version of the pdf file on the same day.

Graham Shepherd

Ericsson "fair co-existence ... in unlicensed band"

Ken Sayers's picture

I recall when I worked for Neighbourhood Cable in the mid 2000's we had a Wi-Fi network in Mildura which was running perfectly until this mob from Tasmania called One Wire (according to Whirlpool they went belly up in Feb. 2008). This mob was sucking money out of the Broadband guarantee program then being run by the Howard Government. They put up a WiFi network which created massive interference to our network. When I spoke to the ACMA I was told "unlicensed spectrum is like the swings in the playground. All the people who want to use the swings have to agree on how they will be shared. There is no police who can come in and set up a regulated system, so unless they are interfering with licensed spectrum, too bad". We ended up having to send our technical guys out to fix their kit so as to reduce interference. So while it might sound like a great idea to use unlicensed spectrum for LTE, that use comes with risks. 5GHz in less risky as it's a pretty broad band, but then it is also pretty short range, so it might be good in a closed space like say a Race Track, as long as the users of that spectrum is limited by e.g. the club that owns the race track.

Ericsson "fair co-existence ... in unlicensed band"

administrator's picture

"Ericsson says ‘The testing validated LTE performance in the unlicensed band and fair co-existence with other technologies like Wi-Fi within the unlicensed 5 GHz band.’"

This is a very big claim. Such matters should be independently determined, not just claimed by a vested interest.

Graham Shepherd

2015 Charles Todd Oration

Douglas OHara's picture

Are there any vacancies for this Sydney event on 26 November 2015?

Regards ... Doug O'Hara

“Just” the Tragedy of Commons: what about LTE-U?

Brian Louey-Gung's picture

Andrew Kerans rightly points out that the WiFi bands are explicitly governed by Class Licenses and that anyone is allowed to play, as long as they abide by the sharing rules. Telstra is using the WiFi standard and as such is following the rules. He also points out quite rightly that at least some of the users being affected by the problem described by Stan Beer will be complicit as they will have prioritised the Telstra service on their smartphones – something that is easily rectified by changing the relevant smartphone options. Maybe the ACCC or the Council of Small Business Australia could create a pamphlet that gives straightforward instructions for the most popular smartphones so that affected businesses could distribute it to their customers.

However, Stan’s article leads to a related WiFi issue - LTE-U. This is where all of the carriers, not just Telstra, will start accessing the WiFi bands to supplement their already massive spectrum holdings. The developers of LTE-U assure us that they are going to great efforts to ensure that it won’t degrade the normal WiFi experience. Forgive me for being a little bit sceptical that their assurances are completely unbiased.

Stan is not the only one to notice WiFi sharing issues. Apparently the Dutch Radio agency Agentschap Telecom published a report recently noting many instances of overloaded WiFi at 2.4Ghz (Sorry, I don’t read Dutch, so I am quoting second hand). So even when everyone plays by the WiFi rules, the Tragedy of Commons occurs – too many users attempting to share a limited resource spoils it for everyone. While opening up the 5Ghz band alleviates the problem, this is just a temporary fix as the user-base grows to fill even this 'enlarged' commons. The deployment of LTE-U will add a massive new user base for WiFi spectrum in one hit.

Rather than relying on the backers of LTE-U telling us not to worry, I’d be much more comfortable if someone like the ACMA analysed the issue and provided independent expert assurances that the ordinary WiFi user is not going to be at the ‘Tragedy’ end of the Tragedy of Commons.

Brian

Is Telstra the bully or is it just the tragedy of the commons

Andrew Kerans's picture

In CommsWire on 7 October Stan Beer decried Telstra for being ‘the bully in the playground’ stealing everybody’s ‘WiFi’ lunch money.  Really, is it that bad and is it just Telstra.

To answer the second question first, no it isn’t just Telstra.  Various technologies use, or are looking at using the ‘free’ 2.4 and 5.8 GHz ‘WiFi’ bands.  They are also perfectly entitled to.  These bands are not actually set aside for WiFi, nor does WiFi have and special rights there.  WiFi must share these bands with whatever else comes along.

The beauty, and ugliness of these bands is that they are practically globally harmonised for the same purpose, Industrial, Scientific and Medical applications, hence the proper term for them, ISM bands.  This global harmony has brought about a proliferation of different ingenious devices doing almost everything you could imagine could be done with radio.  Yet each one of these devices must share the spectrum in accordance with the rules set out in the ACMA Class Licence (in Australia).

Stan also says that patrons of coffee shops ‘find the Telstra service barging its way onto their screens interfering with the weaker local signal’.  Really?  I doubt it.  The powers (in EIRP) are defined in the Class Licence and I doubt very much an organisation like Telstra would flout them the way some other smaller ISPs have and do, particularly when using 5.8 GHz for point to point feeder systems. 

The WiFi radios in the Telstra system would operate to the same rules the smaller WiFi systems do. If the box in the coffee shop is carefully positioned then there would be very little difference in power.  WiFi is cooperative, so just like in a crowded ‘multi system’ shopping centre things will still work OK as long as everybody obeys the rules.  Now if you happen to let your smartphone or tablet ‘prioritise’ Telstra and don’t disable that, then the barging in is really your own fault, albeit a human weakness that perhaps Telstra knows about.

So, the WiFi bands are not the domain of WiFi or of any one or any group of operators.  In my time in the ACMA we called them ‘garbage bands’ because there was simply so much stuff in there.  The other problem with these bands is a soon as one person’s application cops a bit of interference from another they think the ACMA should drop everything, step in and fix it.  Well they simply are not going to, they have bigger fish to fry and besides, playing in the garbage can get smelly.

The value the ISM bands deliver to the community through technical innovation, and yes: WiFi, is undeniable.  But just as the town commons does not belong to any single sheep herder these bands do not belong to WiFi.  So Stan, either pool your lunch money into a communal fund, or find another playground.

Dr Andrew Kerans; Spectrum Management Associates.

IoT security

Tristang's picture

It opens alot of questions, there is alot of of people discussing it. I have organised with Stuart Corner to talk on IoT next month and these will be some good questions to pose to him.

Cheers,
Tristan

Who is researching IoT security?

administrator's picture

Good points, Brian.

Many "things" such as car control systems won't connect directly to the network but, theoretically at least, will be autonomous, using information obtained from radiating and non-radiating devices to inform (or misinform?) them. There will be an eventual network connection point, say, a black box. But what you are saying is that the IoT magnifies the security issue potentially by orders of magnitude.

What work is being done to address this? Will early experiences of insecurity or even the fear of insecurity dampen the takeup of IoT, particularly for vulnerable applications?

Graham Shepherd

Reported and reviewed in CommsWire 15 July 2015

administrator's picture

The Vodafone commissioned report was covered in CommsWire on 15 July 2015 and also reviewed by ECONOMUSE - John de Ridder - in the same issue with some additional strong points made. 

Dr Paul Paterson, head of the Bureau of Communications Research, is speaking at a TelSoc event of 24 July in Sydney. There should be ample opportunity to raise the challenges there. Register here.

Graham Shepherd

'new anti-piracy laws in action'

jonl000's picture

"The first letter sent to purported Internet pirates under the Government’s new anti-piracy legislation..."

I presume this article refers to the leaked draft letter relating to Dallas Buyer's Club and their Federal Court action. This has exactly nothing to do with any government 'anti-piracy legislation'. The government legislation currently before parliament relates to blocking offshore sites.

The sending of letters to ISP customers is the subject of the industry code being developed by the Communications Alliance. This has been undertaken under threat of legislation and with significant pressure from government, but there has been no actual legislation.

The Dallas Buyer's Club case is also happening in advance of that industry code.

This is therefore a remarkably inaccurate headline. Sure, it can be construed as 'bullying', but there's been no change to the law by government that's enabled it.

Net Neutrality

Gary McLaren's picture

Excellent article on the Net Neutrality issue. I particularly like the positioning of Net Neutrality as a public relations issues for the larger ISPs. But the dilemma remains - flat rate plans without download caps encourage higher usage and higher cost but not cost recovery to the ISP. Low usage customers end up subsidising high usage customers and ISPs have no incentive to invest in more network capacity - fibre, routers, transmission systems etc.

Broadband, just like other services where increased usage has increased costs, needs to have a feedback system to encourage investment and competition that prevents dominant players taking monopoly rents.

In theory effective

Graham Shepherd's picture

In theory effective competition would sort this out. Unfortunately, in Australia and in this globalising world, competition is just a theory and nothing more. For some it is a panacea and for others a religion. ACCC plays a key role but it is severely under-resourced and politically influenced.

Trust and rust also play a big part. Telstra hangs on to most of its consumer customers not by price or quality or service but simply because of an ethereal trust factor which might be more likened (lichen?) to rust. It's enterprise customers actually get good package deals but it's poor SME market share shows that discerning customers find somewhere else to go.

One trust factor which Vodafone has identified is that customers have been paying for unused volumes. It goes to the heart of trust. If I put my money in the bank I don't expect the unspent amount to disappear at the end of each month.

I buy a data pack towards the end of the month from Telstra just to get through the month but again I leave a large amount unused and I actually have to take a specific action to turn it off before the end of the month to avoid getting hit again. 

I am sure that Telstra isn't doing this deliberately just ignorantly. Ignorant of the irritation and the growing distrust. The rust starts to shake off.

Overall I think that the market should deal with it but a better resourced and more engaged ACCC might help. I just find it interesting that in their defence companies always resort to talking about their costs when these issues come up and not about their customers. Witness David Thodey and the other telcos arguing about the costs of data retention and not about their customers' privacy.

Graham Shepherd

Cost

Gary McLaren's picture

Well put - but the important thing is that Australia's ISPs have a cost they can't avoid related to network usage. This is going to be a smoother curve (rather than discrete upgrades) in the NBN world due to CVC charges. Telstra's pricing will need to fit into this model - regardless of their content plays. NBN Co will have discrete cost jumps as usage increases - they are small compared to fibre investment but still need to be managed. We all seem to agree that higher speeds should have higher prices - why not higher usage. Usage is a good enough proxy for peak hour dimensioning - just seems natural that higher users should pay rather than be subsidised by lower users. 

What are your thoughts on mobile caps? Should they also be dismantled?

Gary, price and the cost are

Graham Shepherd's picture

Gary, price and the cost are two different things. Generally entirely unrelated. Cost is what you can't avoid, price is what you can achieve, the difference is profit (or loss). It's not clear to me that what Telstra charges are in anyway representative of actual costs, particularly given their extraordinarily high gross margins and overheads. Peak hour volumes determine dimensioning of networks but costs actually rise in steps - a large step for a major infrastructure build, eg, a cable, and then smaller steps for lighting up a fibre and adding wavelengths. Likewise routers, DSLAMS etc. In HFC the DOCSIS upgrade costs are actually quite low - the biggest steps to achieve real speeds are deeper fibre as you know and these are speed related not necessarily volume related.

Monthly pricing by volume might also be seen as inequitable. Every month most consumers use significantly less than their quota. Vodafone's initiative to rollover unused volume is a welcome introduction to the Australian market and perhaps reveals something more of costs and competitive pricing.

As Telstra gets  more and more into content it will have the same motivation as others to eliminate metering of their own or affiliated content and the relative cost to them will be acceptable in their business case. 

I suspect that Australia's access pricing structure is based on history (infrastructure monopoly) and opportunity. 

Effective competition in access infrastructure, as you have often said yourself, would achieve the best outcome in the long term (coupled with an effective USO).

Graham Shepherd

Australian backhaul costs

Gary McLaren's picture

Graham your comment on backhaul is interesting. Costs to the US for Australian major ISPs with scale are now pretty low - appox. $10 per Mbps per month. Costs around metropolitan Australia are lower again - regional Australia is another matter because of limited competition to NBN Co's regional PoIs.

Telstra and NBN Co charge VC costs  - $17 to $30 per Mbps. So all up costs per Mbps are around $30 to $50 per Mbps. So someone who downloads 200Gbyte (roughly equivalent to 1Mbps for a month assuming some time for sleeping!) costs the ISPs about the same - $30 to $50 for a 200Gb plan

So this is a significant cost still that can't be absorbed or subsidised by the low end user.

As you can see the main cost going forward will not be domestic and international backhaul but the VC charges of Telstra and NBN Co. NBN Co's will drop over time as volume increases.

But a price signal is nevertheless important - no one has given me an argument why low volume users sshould ubsidise high users? I get it in healthcare for good social and economic reasons - a healthy population is good for all - but not aware of any other business model that is sustainable on this basis.

This is a bit of politically

Graham Shepherd's picture
  1. This is a bit of politically correct hyperbole about NBN. In its present configuration both structurally and technically it will never achieve 1Gb/s for consumers.
  2. I suspect that the real problem for data caps in Australia is backhaul. Australian backhaul is being addressed but backhaul to the US  and Europe is still far too expensive. Otherwise I agree with Hastings that there is otherwise no real need for data caps.
  3. Ubiquity of availability and equitable low pricing is a far better way of adddressing piracy than the draconian laws now being put forward in Australia and by the TPPA and TAFTA. Low pricing has mostly to do with cutting out the middle men. Carve a Suez Canal through them and you have solved most of the problem. 

Graham Shepherd

Lost NBN Opportunity

Robert Reid's picture

Let's keep this NBN discussion simple by making a few comparisons.  Australia is operating beyond capacity on our roads so decision makers arrange for freeways to enable greater traffic capacity.  While Rail transport has even greater efficiency is seen as costly to implement just like the original NBN it had to be developed.   While road and rail have a national distribution to keep pace with society needs they at times need new branching infrastructure to reach customers unlike Australia’s telecoms customer access network that is outdated in terms of future capacity and performance and is costly to maintain.  

While there is diversity in the road and rail networks they are not actually duplicated or indeed exclusive allowing freedom of choice for the end user compared to what is now proposed for the NBN.  No one would consider duplicating roads and having different roadway service providers as with the other utilities of power, gas and water!  These other utilities relate to the NBN very well as every house or business is accessed by a single street “an address” , just like individual telecom services i.e. the humble phone number, and interconnects ultimately to multi user infrastructure of choice - the core road network owned by the community for all to use. 

The original and much considered fibre based NBN was conceived to provide a very high capacity single point of entry passive connection, offering a high upside, to unique street addresses just like the old copper network.  The NBN offered the promise of very low maintenance compared to the copper network and indeed any of the proposed networks of FTTN, HFC etc.  Compared with land owner a street address, the freedom of choice is based on the land owner’s choice of where to reside not the actual roadway, water or power infrastructure.  Indeed at the extreme the land owner could opt for no services at all with bush track access!  Clearly for the future modern society telecommunications and its performance has become a much needed utility and needs to be delivered efficiently at the lowest possible cost over time bearing in mind future needs.

While the original fibre based NBN appeared to be a higher cost with a slower rollout, and an issue not well understood by the community at large, this was due to it being an entirely a new network that took account of likely the future needs of all Australian’s, as was considered in the mid 1950’s with the now outdated copper network that in relative terms cost much more to a considerably smaller community of Australian’s.  Imagine where Australia would be now if it wasn’t for this level of forethought and investment.  With this in mind, quality NBN fibre based infrastructure, while initially more costly, ultimately offers a lower cost centrally managed operational technology with an almost indefinite data capacity upside compared to any other medium.  The NBN as originally proposed was competitive as it provided freedom of service and service provider as by way of comparison is provided by our road, water, power, gas infrastructure.  Any alternate approach significantly increases the long term operational costs [OPX] that is as yet un-costed and CAPX due to short infrastructure life due to the need to service nodes and copper tails with almost no upside capacity as it is finite compared to the original NBN FTTH. 

The outcome of what is now set to become the mixed technology and service provider delivered NBN will incorporate unnecessary complications to the community under the guise of competition due to the mix of network operators who under the present deregulated framework while the user pays all the hidden costs that that appear not accounted for by the decision makers.  Of question in respect to the proposed new NBN - do Australians really want or need competition like they have with the power industry that few if any in the community understand?  Indeed any comparison with the power retail network providers appears to indicate higher cost while supposed competition is keeping prices low! 

As the power and water distribution networks are not duplicated as is intended in the NBN.  This begs the question for all Australians – why is:

the national communications network being nobbled to limit it to short term technology that will become obsolete with inbuilt un-costed OPX?

a network requiring higher maintenance and power usage than that originally intended being proposed? 

is competition being argued for the telecom customer access and distribution network when just like cars on roads [water, power etc.] the end user only requires choice in respect to their end product not how it is delivered?

Australia appears to have been be hijacked by a lack political vision clouded by supposed open market forces suggestive of open competition benefits without an appreciation of the economics of scale such as all other utilities of necessity use.  To draw a comparison, it would be seen as ridiculous to build separate roadway infrastructure layered above or below each present road to allow freedom of choice of roadway service provider in order to drive to a destination.  Imagine the complications of a service provider to manage this costly competition model and yet this is what the proposed new NBN.  This approach, like telecoms, unnecessarily implements a system of tole operators to deliver their services on their network as well as those of their competitors.  How will USO manage this and if so how and who is to arbitrate on the issues the TIO?  Why make a simple network requirement so complex – surely Australian citizens don’t want or need this in the name of competition.  This begs the question - how is the service level to be managed on these new NBN network islands and indeed between these network islands - by the end customer???  This is farcical as it provides services that are ultimately higher in cost and have a shorter technological shelf life while competition is actually compromised not enhanced – can Australia afford this? 

To summarise, just like our national road and rail networks and the water, power and gas delivery infrastructure, the lowest cost ultimately can only be delivered by a well thought out and implemented service delivery network - not several separately managed networks as is proposed for the new NBN.  Why is it that the NBN is seen as so different to other centrally managed infrastructure?  Everyone knows that Australia will ultimately require FTTH as is delivered now to big business and even the telecom service providers for their core networks, so why are the Australia's decision makers deliberately taking the more costly route?  What is proposed appears short term, ultimately more costly and will delay high capacity broad band to domestic, small business Australians and all forms of end needs such as medical and road management and is seen as having has nothing to do with real service driven competition.  Let’s see some real analysis and justification rather than political idealism under the guise of competition.

HFC Exploitation...

Craig Watkins's picture

I appreciated many great points both in the original article by Gary, and in the above comment by Graham. Ultimately we must be fully aware of the long-term trends in video content that dictate directions that our communications systems take.

Globally the cable industry has become an important player in the provision of broadband services. There is even reason to suggest it has a valuable role already in the Australian context, despite comparably low penetration rates. In the US market it is easy to see how competitive forces coupled with high Cable TV take up rates from 20+ years ago have led to strong cable broadband offerings today.

The outlook is clearly for access technology of any flavour to be the means for delivering IP broadband service, and other video content services being over the top of this. Of course the larger companies can be expected to flex their 'monopoly' power to the extent that this is allowed. Some companies will certainly attempt to lock up popular content and maintain regular monthly subscription services. However, many consumers are also likely to source content from a large range of disparate sources. The popularity of YouTube is unlikely to wane in the near future, and there is additionally an enormous wealth of "special interest" content that exists beyond YouTube.

The issue with HFC content delivery would seem to be ensuring that the access network component is up to the task of this continued rapid increase in throughput. The headline speeds of 100 Mbps is obviously not the problem. The amount of sharing on DOCSIS nodes is the concern. If the HFC NBN lives up to any sensible expectations, then we must assume that a significant number of new cable lead-ins will need to be constructed. There is a cost associated with these lead-ins. Bill Morrow when recently speaking with Phil Dobbie, has suggested that for aerial lead-ins the cost is tiny. Whatever the reality is on that score, the fact is that the cost of such lead-ins extends beyond the lead-in itself. With the addition of more premises on any physical HFC node area, there quickly becomes a need to split nodes (in excess of that which might be necessary only due to organic demand growth).

We have not yet seen any detailed discussion of HFC and these types of issues from NBN Co. With luck we will see something before long (it is almost 12 months since the Strategic Review document was published). The concern is obviously, as Graham has suggested, that the real cost of HFC use as part of the NBN is likely to be significantly higher than indicated as part of the Strategic Review.

What is the alternative? Is it economic to exploit both the HFC infrastructure and the twisted pair copper infrastructure in the same area? A priori this sounds like a bit of a nightmare, locking in duplicate and disparate infrastructure forever. The rationality of FTTN in 2014 is not particularly compelling for non-HFC areas if you contemplate sensible demand growth projections. FTTN in contrast to HFC offerings would seem even less appealing. If small nodes were used (maximum of a couple of hundred meters from premises), then the performance offered by Vplus (recently announced by Alcatel-Lucent) represents close to the maximum we can expect. This might be a good level of capability for today, but the picture is unlikely to be quite as bright in even 5 years.

A perhaps similar investment compared to the small node Vplus option, is likely to deliver a significant number of HFC aerial cable lead-ins and associated node splitting. The questions here would seem to be what size of HFC node is going to serve the longer-term data throughput requirements, and how many rounds of node-splitting are needed to get there (over what period)? Certainly the economics of having trucks and crews in the same area two years after they were first there would seem a bit foolish. On the other hand a single-step upgrade to the 'end game' is likely to be much higher initial investment than prudent.

It seems that to properly exploit the HFC infrastructure we need solid answers to these sorts of questions. The chance of us obtaining sensible answers is not strong if we follow flawed demand growth logic as the Communications Chambers report appears to have done in relation to video compression technology misunderstandings. In any event, there have been a lot of clock ticks since 2009, and many in the last year. We need to see some evidence that there is a sensible plan for the HFC infrastructure.

The cost-benefit equation of HFC investment should be compared to a baseline of deep-fibre FTTdp, as this promises network capability effectively identical to full FTTP (fibre drops able to be added on demand by individual premises as required in the future, and very high DSL-based connectivity today), at a reduced cost by avoiding the expensive fibre drops. For aerial fibre distribution the FTTdp approach is unlikely to offer a positive cost-benefit equation to full FTTP, so this might be expected to be the baseline comparison there. The point to having a systematic baseline is that we can properly assess HFC investment opportunities and make rational determinations on the optimal way forward. HFC clearly offers the ability for the NBN to cobble something together quickly (and perhaps cheaply depending on negotiations with asset owners). The option must be properly assessed in comparison to the long-term evolution plan and the total investment of cost and time to deliver that over HFC versus other baseline options.

I know there are some really great minds at NBN Co, but it does concern me that we see little evidence of progress. A number of comments made by senior executives in Senate hearings appear to strengthen such concerns. The nation needs a first-class NBN outcome. Most importantly we need to know we are heading towards such an outcome. 

Pages