ACCC Can solve 'Crazy' Rural Broadband problems
This is an executive summary of a longer letter written by the author to the editor of Communications Day, in response to an article about the costs to create a new national regional broadband infrastructure. It explains another approach that would yield competitive regional broadband at very little cost, across the nation..
Mr. Hackett, the managing director of national broadband innovator Internode and Internet age telco Agile Communications, said it was "crazy" to suggest that the Government should build new local loop networks in rural towns to deliver high-speed broadband. "It's now viable to construct new regional local loop networks in country towns, but the one true economic barrier to 'bush broadband' is the cost of backhaul back to the big smoke," said Mr. Hackett.
"The Telstra fees to acquire intra-state backhaul from a regional or remote township remain one of the last bastions of 1980's monopoly pricing. The current cost-per-megabit-per-month from Telstra for intra-state backhaul circuits is 20-100 times too high to make those circuits viable for broadband services provision by competitors."
"There's a monopoly in access to long haul regional fibre runs intra-state to thousands of small regional townships across the country. It's largely a hidden monopoly because its role in wrecking the economics of competitive regional broadband is largely unsung."
"The solution is deceptively simple: It just needs the ACCC to declare access to Telstra long haul regional fibre into Telstra exchanges - to simply declare that the Line Sharing Service also applies to monopoly fibre runs, not just monopoly copper ones."
Internode (www.internode.on.net) is a leading Australian broadband services and Internet services provider. Its wholesale partner, Agile Communications, is an Internet age telco that uses innovative packet-switching technology to achieve price-performance points unavailable through old-world, circuit-switched carrier networks.
During the past four years, Agile has built a microwave-based long haul backbone network into the Coorong region of South Australia. The company overbuilt this network because Telstra's intra-state backhaul costs were too high, a situation that remains unchanged.
Mr. Hackett made his comments in response to the recent proposal by National Party "Think Tank", the Page Research Centre, to spend $7 billion from the proposed sale of Telstra to open up regional telecommunications outcomes with high speed broadband networks.
"The crazy thing about such a notion is that neither the market nor its customers need the government to create new local loops," said Mr. Hackett. "Existing copper loops, with a dose of upgrade here and there, can do that job today. A 24 megabit ADSL2+ service can carry multiple high definition TV channels and still have 10 megabits left over for Internet access. That bush broadband problem is on a clear path to solution already, powered by the one enabler it ever needed - the declared service access to Telstra copper loops."
Mr. Hackett said line-sharing of fibre cables is possible through Wave Division Multiplexing (WDM) which literally shares access to a single fibre pair. "It shares it in a manner that allows an arbitrary number of competitors to access the same fibre pair, with no detriment to the original use of that fibre by Telstra," he said. "Multiples of 1 Gigabit per second or 10 Gigabits per second, on each wavelength, delivered via a low cost WDM device, would create more contestable bandwidth into a regional town than it would ever need."
"This will drive change because Telstra only lowers service pricing due to competition (in contestable markets) or due to ACCC intervention (in monopoly markets). Intra-state backhaul is a monopoly market. The required intervention is simple and the community benefits would be massive."
(edited version of) A letter to the Editor of Commsday, from Simon Hackett - M.D. Internode and Agile Communications
I read, with interest, your commentary on the Page Research Centre's concept to create, in effect, "GovTelco" for the princely sum of $30 billion or perhaps $7 billion or whatever to open up regional telecommunications outcomes (CD, 18 March).
The crazy thing about such a notion is that neither the market nor its customers need the government to create new local loops.
Existing copper loops, with a dose of upgrade here and there, can do that job today.
With a 24 megabit ADSL2+ service able to carry multiple high definition TV channels and still have ten megabits left over for Internet access, that problem is on a clear path to solution already, powered by the one enabler it ever needed - the declared service access to Telstra copper loops.
One look at FastWeb in Italy demonstrates how effectively this nails the practical issues in the local loop.
In the presence of that declaration, which, despite Telstra efforts to the contrary, is working, and will keep working, and in the light of the Australian government HiBIS scheme, it is now viable to construct new regional local loop networks in country towns.
But what HiBIS cannot solve, in practice, is the one true economic barrier to bush broadband, and that is the cost of backhaul back to the big smoke.
The Telstra fees to acquire intra-state backhaul from a regional or remote township remain one of the last bastions of 1980's monopoly pricing (the other one is the price of inter-capital transmission to Hobart).
The reason for this is simple and obvious.
Telstra only lowers service pricing in the presence of competition.
In any market in which Telstra is the only provider, pricing simply remains immovable.
The current cost-per-megabit-per-month from Telstra for intra-state backhaul circuits is as much as 50-100 times more expensive than is needed to make those circuits viable for broadband services provision by competitors.
However, if Telstra decides to create broadband access, it charges itself such a small amount for backhaul that it can create regional pricing on a basis impossible for its competitors to match.
This, of course, is the definition of monopoly, and monopoly rent.
Backhaul Costs Prohibitive:
So, let's examine those regional townships and the cost of that backhaul.
Intra-state backhaul is also delivered, in almost all cases, over existing Telstra optical fibre.
The old Telstra microwave tower chains in the bush are still there too, but largely because it is more expensive to pull them down than to leave them up.
I learned a salutary lesson some four years ago.
My company had just built a microwave-based long haul backbone network into the Coorong region in South Australia.
It extends from Adelaide out as far as a small country town called Tintinara, a whistle-stop on the way to Melbourne.
We had overbuilt this microwave network precisely because the Telstra intra-state costs were too high.
They have remained basically unchanged since.
I was taken on a Telstra exchange tour in Tintinara, notable for two amazing reasons.
The first was the reality of how small the required telephony equipment had become - the exchange building had been emptied of the first two generations of voice switching gear, leaving an empty, private hall large enough to play touch football in, with the current generation of voice switching gear now so small that it had been relocated to the toilet next door (literally!) because that room was smaller and hence cheaper to air condition.
But the second revelation was the deep one.
We had spent 12 months and millions of dollars creating a new long haul microwave backbone to Tintinara from Adelaide, quite simply because the Telstra charges for intra-state bandwidth were too high to make them viable for broadband use.
And yet there in the corner of this old building, in the middle of nowhere, was the Telstra optical fibre, coming into the building and plugged into that switchgear in the toilet.
This optical fibre was capable of carrying all of the phone calls in Australia, at once, as Telstra brochures used to cheerily remind us in the 1980's.
Or in modern terms, tens of gigabits per second of data traffic.
And there it was, servicing a tiny a township with a few hundred people and a Remote Switching Stage in the loo.
The visible and accelerating revolution in competitive metropolitan broadband in Australia is happening for one reason, and one reason only - the ACCC decision to declare access to Telstra monopoly copper lines in Line Sharing Service mode - to insist that access be granted to the high frequencies on copper voice lines at a set cost, under defined access conditions.
As a direct result of that declaration, we now have many competitors, ourselves included, delivering ADSL services that are today operating at peak speeds above those of Telstra Cable Internet services.
And where we can manage to create or access backhaul, we are doing it in country towns as well.
Two townships attached to our microwave backbone already have 12 megabit per second ADSL2 broadband, today, and Tintinara is next on the list.
But what about the thousands of other towns, where no one has managed to somehow fund a new backbone network to make the backhaul costs viable?
The Simple Answer:
Now we come to the punch line.
How would we enable competitive provision of regional broadband and fast?
How could we do it without spending $7 billion on creating "GovTelco" and having it ultimately implode under its own administrative weight, while taking 10 years to recreate a massive fibre backhaul infrastructure that already exists?
It's deceptively simple: All it would take would be for the ACCC to declare access to Telstra long haul regional fibre into Telstra exchanges - to simply declare that the Line Sharing Service also applies to monopoly fibre runs, not just monopoly copper ones.
To be clear, I am not referring to fibre in metro areas.
That is an area where there is plenty of competition, for the right reasons, and there is no access problem to solve.
Neither is there an interstate fibre access monopoly.
To whatever extent Optus started to fix that problem with their inter-capital fibre build, NextGen definitely solved it - to the benefit of us all.
And the pricing for inter-capital transmission (including Telstra's offerings) has tumbled down to viable levels as a result.
But there is a monopoly in access to long haul regional fibre runs intra-state.
In fibre runs to Tintinara and to two thousand other small regional townships across the country.
It is largely a hidden monopoly because its role in wrecking the economics of competitive regional broadband is largely unsung.
So unsung, that not even the Nationals' think-tank, the Page Research Centre, realises that the solution to its perceived problem already exists and does not need to be built again (exactly as for the copper local loop network).
So how would line sharing work when the line is made of fibre?
Telstra would surely argue that there are no spare fibres left.
The answer comes from the glossy brochures of Telstra and a thousand telco equipment suppliers across the world: Wave Division Multiplexing (WDM).
WDM, literally, shares access to a single fibre pair.
It shares it in a manner that allows not just a single competitor to access it concurrently (as for copper-based LSS).
It allows an arbitrary number of competitors to access the same fibre pair, with no detriment to the original use of that fibre by Telstra.
More Bandwidth Than Needed:
Multiples of 1 Gigabit per second or 10 Gigabits per second, on each wavelength, delivered via a low cost WDM device, would create more contestable bandwidth into a regional town than it would ever need.
Exactly as for copper LSS, the operating and maintenance cost of that Telstra fibre run is fully funded by the existing use of the fibre to support voice telephony and Telstra data services into these townships - or it would not already exist.
So the costs for fibre LSS access, like those for copper LSS, can and should be confined to the physical costs of installing WDM equipment (and borne by access seekers), and to a similar monthly access cost for each use of that fibre to that imposed for copper LSS (circa $9 per month per path) for the costs to Telstra of sending out a monthly access bill.
Competitors would happily fund the WDM equipment, and the basis for doing so could be the same as the basis on which Telstra currently manages and operates the process where competitors fund equipment space in exchanges and subsequent competitors then gain access to the same rooms on a shared cost basis.
So. It's this simple:
- Declare access to intra-state monopoly fibre paths, using a declaration based on the existing copper LSS Declaration
- Fund WDM equipment for multiple-competitor access on the same cost sharing arrangements as used today for build cost sharing
- Set the price based on the copper LSS costs
- Stand back and let competitors get on with it, using HiBIS to co-fund the deployment of DSLAMs and wireless systems in the regional towns concerned.
Total additional cost to the government?
Rather than spending millions to recreate a national regional fibre network, the costs of this alternative approach boil down to ACCC time and resources to make and enforce the Declaration.
Alternatively, regulatory intervention in respect of the existing price of 'wholesale transmission' services from Telstra (which are already a Declared service) could potentially achieve the same effect.
While backhaul using this Declared 'Transmission' service is available today, the current pricing offered by Telstra is up to one hundred times more expensive, per megabit, than would be required to allow sustainable regional broadband to be constructed.
That existing pricing makes it impossible to create regional broadband services that are price competitive with Telstra ADSL services. So today, if Telstra don't think its worth installing broadband in your town, you just don't get it.
Doing the sums on the existing Wholesale ADSL service fees from Telstra, its clear that the internal 'transfer cost' Telstra uses to price backhaul when it provides ADSL is massively lower than the price it charges for the backhaul component to those same towns, to others. It uses a completely different pricing structure.
With these regional backhaul prices re-framed (or shared Fibre access Declared) to level the playing field, all carriers would have equal access to the irreplaceable, existing, regional backhaul routes at the same true (low) price Telstra charges itself for that access. As a result, in most country towns, we would see multiple regional infrastructure builders providing city-priced, sustainable, broadband using a variety of appropriate local-loop technologies.
Within five to 10 years, most country towns in Australia could have city-quality or better broadband services, funded by a combination of HiBIS subsidy and direct customer fees, and based on whatever technology that regional telcos find to be the most effective in a given township.
The presence of viable and healthy competition would ensure, naturally, that customer costs would be, and would remain, competitive, because the barriers to entry in those towns for subsequent competitors would be low.
If, as in Tintinara, today, Telstra is not offering broadband in a given country town, what possible rational argument exists for denying fibre LSS as an approach to allow a competitor to do it instead?
Oh, yes, there is one argument...
But would the Telstra share price be diminished significantly by allowing Fibre LSS in regional townships?
I would argue that the reverse would be true.
Telstra will be so busy charging competitors access fees for the use of its copper in thousands of townships across Australia that it might actually do better than it does now and without having to do any work, at all.
And that $30 billion could be spent on hospitals and schools - where it should be spent.
Indeed, all those hospitals and schools, in the smallest of country towns, would be able to access affordable broadband at city prices.
Sounds like a plan, doesn't it?
Simon Hackett MD, Agile Communications, Internode