Chapter Two
The new proposal
2.1
The Panel of Experts released its Evaluation Report on the outcomes of
the government's Request for Proposals (RFP) to Roll-out and Operate a
National Broadband Network for Australia on 21 January 2009. Following
the release of the report, the government repeatedly stated that it was
considering the report, but failed to make any announcement on the outcome of
the bids. This lead to months of industry speculation and a high level of
uncertainty within the telecommunications sector.
2.2
On 7 April 2009, the Rudd Government made a joint ministerial
announcement of the 'establishment of a new company to build and operate a new
super fast National Broadband Network.'[1]
The announcement signalled the termination of the RFP process, based on the
view of the Panel of Experts that none of the national proposals offered value
for money, which was the overarching qualifier upon which all RFP evaluation
criteria were based.
The 'New National Broadband Network'
2.3
The announcement detailed the government's new policy direction and included
a commitment to ensure that 'every house, school and business in Australia will
get access to affordable, fast broadband.'[2]
Specifications of the new network
2.4
Although highly anticipated, it would be fair to comment that the
details of the government's announcement took most in the industry by surprise.
While the previous RFP provided the option for proponents to utilise either
fibre-to-the-node (FTTN) or fibre-to-the-premises (FTTP) network architecture,
there was no such option within the new proposal. In addition to specifying
that the technology would be FTTP, the network also was to support a large
increase in the speed of broadband services.
2.5
The new National Broadband Network (NBN) was to:
-
Connect 90 per cent of all Australian homes, schools and
workplaces with broadband services with speeds up to 100 megabits per second;
-
Connect all other premises in Australia with next generation
wireless and satellite technologies that will deliver broadband speeds of 12
megabits per second; and
-
Directly support up to 25,000 local jobs every year, on average,
over the eight year life of the project.[3]
2.6
The announcement stated that the FTTP network will extend 'to towns with
populations of around 1000 or more people.'[4]
The fact that the remainder of the population not reached by fibre would now be
guaranteed the same minimum speed promised to metropolitan subscribers under
the previous RFP process, could be seen as a benefit to those living in remote
areas.
2.7
In addition, the government announced its intention to provide new fibre
optic transmission links connecting cities with major regional centres and
rural towns. This measure addresses the issue raised by many within the
industry that the lack of backhaul access and interconnection is a major factor
in the dearth of affordable broadband in areas of lower population densities.
2.8
Most telcos welcomed the new proposal, as was reported widely in the
media in the days following the announcement. Communications Day
provided a concise sample of industry commentary, which included statements
from iiNet, Optus, Primus, Internode and Macquarie Telecom. For example, Mr
Maha Krishnapillai from Optus was reported as saying that the government had taken
'a visionary and nation building step in the right direction', while Mr Michael
Malone from iiNet reportedly said:
This is the best of all possible outcomes ... In terms of key
criteria we were looking for in a National Broadband Network – open access,
structural reform, fixing backhaul 'blackspots' and regulatory reform – the
government has delivered.[5]
2.9
A main feature of the new network was the announcement that it would be a
national wholesale-only, open access broadband network. This sent a clear
message to the telecommunications industry that the government intended to
impose strict regulatory reforms in order to address competition issues in the
current market.
2.10
As a supplementary feature, the government announced that the building
of the NBN was to be a 'major nation-building project' with the ability to
support, on average, 25,000 local jobs every year, a figure that would peak at
37,000. This announcement was welcomed in view of the impact of the global financial
crisis across many Australian sectors. The government also claimed that not
only would this major infrastructure project stimulate employment in the short
term, it would also provide productivity gains and increased innovation, the
benefits of which would 'continue to flow for decades beyond the completion of
the project.'[6]
NBN Co Limited
2.11
A major aspect of the project was the establishment of a new company to
build and operate the new network. This company has since been registered as
NBN Co Limited (NBN Co). While the government is listed as the company's major
shareholder, 'significant private investment in the company is anticipated'.
The government has committed to an initial investment of $4.7 billion, with
joint private sector investment of up to $43 billion over the build time of 8 years.[7]
2.12
The government will seek private investment of up to 49 per cent of the
company, with the objective of benefiting from private sector capacity and
expertise in the telecommunications industry. However, there are to be
limitations on ownership to ensure the government can deliver on its promise of
retaining the network as a wholesale, open access operation.
2.13
The government has stated its intention that, once fully operational, it
will sell its share in NBN Co to the private sector:
The government will make an initial investment in this
company but intends to sell down its interests in the company within 5 years
after the network is built and fully operational, consistent with market
conditions, and national identity security considerations.[8]
2.14
There is speculation that NBN Co as a regulated monopoly provider will leave
the Australian telecommunications sector in a similarly uncompetitive position
to that which the government is currently seeking to address. Until full
details of NBN Co's governance framework are made available, including any
ownership limitations, this speculation will undoubtedly continue.
2.15
The government has stated that its investment in the company will be
funded through the Building Australia Fund (BAF) and the issuance of Aussie
Infrastructure Bonds (AIBs), providing an opportunity for households and
institutions to invest in the NBN. Further details relating to the funding of
the NBN can be found in chapter five.
2.16
Further details of the operation and funding of the NBN Co are also
discussed in chapter five.
Plan of action
2.17
The government has also announced a 'plan of action'[9]
to launch NBN activities, stating that they would immediately:
-
Commence an implementation study to determine the operating
arrangements, detailed network design, ways to attract private sector
investment – for roll-out in early 2010, and ways to provide procurement
opportunities for local businesses;
-
Fast track negotiations with the Tasmanian Government, as recommended
by the Panel of Experts, to build upon its NBN proposal and begin the roll-out
of a FTTP network and next generation wireless services in Tasmania as early as
July [2009];
-
Implement measures to address 'black spots' through timely
rollout of fibre optic transmission links connecting cities, major regional
centres and rural towns – delivering improvements to telecommunication services
in the short term;
-
Progress legislative changes that will govern NBN Co and
facilitate the rollout of fibre networks, including requiring greenfields
developments to use FTTP technology from 1 July 2010;
-
Make an initial investment in the network of $4.7 billion; and
-
Commence a consultative process on necessary changes to the
existing telecommunications regulatory regime.
Closer examination of detail
2.18
The committee shared the surprise expressed by many within the industry
at the announcement of the new NBN proposal. The new NBN amounts to a major
shift in government policy, requiring architecture delivering FTTP to 90 per cent
of Australian homes, schools and businesses, with alternative technologies of
satellite and wireless proposed for more remote communities.
2.19
In its first interim report published in December 2008, the committee
concluded that the then NBN platform 'should be broadened to enable a greater
level of technology convergence where more appropriate than fibre.'[10]
Consequently the committee acknowledges this aspect of the announcement as a
welcome improvement.
2.20
The committee believes that the decision by the government to nominate
FTTP architecture over the previous, optional FTTN architecture, reflects the general
consensus expressed by key industry stakeholders that investing in FTTN would
result in a network based on out-dated architecture that would not be future-proofed.
FTTP vs FTTN
2.21
Indeed, the government's change in policy direction is reflective of evidence
taken by this committee throughout the RFP process period, highlighting that a FTTN
network could not subsequently be efficiently and effectively upgraded to FTTP.
Representative of this view was Dr Ross Kelso, when he stated that:
I am particularly concerned about prescription of fibre to
the node technology for the national broadband network. I believe that ... if we
are to move down the path of the network being engineered for fibre to the node
where it makes it difficult for it to go beyond that to fibre to the home, it
is a retrograde step.[11]
2.22
There is general consensus throughout the telecommunications industry
that FTTP architecture is the only option that will support future technology
upgrades, given the rapid changes in telecommunications technology, even over
the last five years.
2.23
Broadband is currently being delivered to many Australian homes through
ADSL technology, which involves specialised modems situated in telephone
exchanges utilising existing copper wire networks. The problem with ADSL is
that, due to the limitations of the copper infrastructure, the maximum speed
that data can reach is limited by the home's distance from the exchange. In
fact ADSL will not run effectively beyond certain distances, with around four
kilometres generally accepted as the limit.
2.24
FTTN resolves this problem by decreasing the distance that the data
needs to travel over copper. The FTTN proposal would have seen fibre rolled
out, generally from the local telecommunications exchange, to a 'node' or mini-exchange
that would be located on the footpath. These cabinets, generally within 800m of
a consumer's premises, would house the DSLAM equipment of numerous service
carriers that was positioned in the exchanges. They would also need to be large
enough to house the air conditioning equipment required to keep the DSLAMs at
the correct operating temperatures.
2.25
By taking the fibre closer to the premises, FTTN would decrease the
distance impediment experienced by ADSL technology, whereby the further a
customer was from the exchange (or DSLAM equipment), the greater the likelihood
that the customer could not access ADSL. Although FTTN would have been a
significant step forward, it did not eliminate the dependence on copper
infrastructure, and the associated problems of age deterioration and also the restrictions
inherent in the amount of data even new copper can carry.
2.26
FTTP eliminates the dependence on copper. Each premises will have its
own optical fibre connection from the street to the outside of the premises,
with a connection to a new type of modem that is capable of converting the
optical signals. Fibre optic cables are composed of strands of pure glass, the
dimensions of human hair, which carry data over long distances in pulses of
light. Because data is transmitted in light pulses, distance no longer impacts
performance, so that a premises located 30 km from the exchange will receive
data at the same speed as a premises right next to the exchange.
Technology explained
2.27
There are two main options for the government to consider when choosing
the FTTP technology: point-to-point (P2P) or Gigabit Passive Optical Networks (GPON).
The government has stated that it will use leading edge technology in the
deployment of the FTTP network, but has declined to provide more specific
details, stating that this level of detail will be provided in the
Implementation Study, due by the end of February 2010. However, in responses to
Questions on Notice taken during May Budget Estimates, the Department of
Finance and Deregulation stated that:
DBCDE considered that for the local distribution component of
the FTTH [FTTP] network that a passive optical network was the most appropriate
basis for the development of a preliminary cost estimate.[14]
P2P
2.28
Point-to-point technology would see every premises allocated a dedicated
fibre, which would connect to a local Optical Line Termination (OLT). These
OLTs would need to be located on most street corners, in a similar fashion to
the 'nodes' under the previous FTTN proposal. OLT cabinets would need to
contain significant electronics and would require cooling in hot weather.
2.29
P2P may seem ideal in providing dedicated fibre to every customer, thus
providing greater scope for service differentiation. However, in reality this
option would be far more costly to deploy and would also result in street-scape
issues and noise pollution from the electronics and air-conditioning within
each OLT.
GPON
2.30
In the GPON alternative, a single optical fibre is utilised for multiple
premises, which then share the bandwidth available on the fibre. As explained
in an Alcatel-Lucent brochure:
In a GPON environment, a single fibre runs from a central OLT
site serving up to 64 users. Consumers up to 30 kilometres away can be
economically connected on this single fibre. Close to the consumer's premises,
the cable is split inside a junction box, similar to those used in today's
telephony network. ... No power is required at any point between the exchange and
the home installation.[15]
2.31
The GPON option would be more cost efficient, due not only to the
reduced amount of fibre required but also the corresponding reduction in the
number of fibre joins. This naturally translates to less man hours and labour
costs for a GPON deployment. Another advantage of GPON is the fact that it
requires no power between the exchange and the premises. As the brochure
concludes:
...it is probably fair to say, from a visual, noise and carbon
footprint standpoint, GPON is preferable for residential fibre coverage.[16]
Possible functional layers
2.32
The NBN is underpinned by the government's policy requirement that it
will be an open access, wholesale-only network. Although to date the government
has not elaborated on how this will operate, some suggestions have been made by
the industry.
2.33
It is reasonable to assume that there will be three basic types of
service providers, with three corresponding functional layers of the network,
as follows:
-
NBN Co: The public-private partnership established
to build and operate the NBN. Returns are assumed to be regulated and the
company will be excluded from providing retail services. This company provides
wholesale access services to the Network Service Providers.
-
Network Service Provider (NSP): NSPs will have a
retail relationship with customers and provide Internet protocol (IP) access to
applications. They may choose to develop and provide applications themselves.
-
Application Service Provider (ASP): ASPs provide the
applications such as television, video, voice telephony and internet access.
They will also be the providers of applications and services that are yet to
emerge from non-telecom areas such as health, education and power management.[17]
2.34
Today, a typical Internet Service Provider (ISP) fulfils the role of an
integrated NSP/ASP, having retail relationships with customers and providing
access to standard internet services, while others also offer telephony and paid
video services.[18]
Network architecture
2.35
From this point the options multiply almost exponentially, with
considerations before the Implementation Study that will include: connectivity
for multi-dwelling units; the Optical Network Termination (ONT) device and its
connection to one or more Residential Gateways (RG) within the premises; ownership
of those devices; battery back-up requirements; and billing options.
2.36
One of the many complex decisions required will be the determination of
the wholesale point of interconnect and service boundary point scenarios. A
useful diagram illustrating the end-to-end architecture vision was provided in
Alcatel-Lucent's submission, and is copied at Figure 1 below.
Figure
1[19]
The 90/10 footprint
2.37
Australia possesses a geographically diverse topography, with a dense
population around the coastline and sparse, but often economically significant,
communities scattered across remote areas. After terminating the FTTN proposal,
the government has included 'next generation wireless and satellite
technologies ... to people living in more remote parts of rural Australia.'[20]
2.38
The committee notes that there is still varied opinion as to whether the
policy will result in what could be seen as a broadband 'patchwork' rather than
a network, and how the subsequent risks regarding the capability for national
connectivity can be mitigated. This will need to be resolved in the context of
the network architecture solution, which should be a major component of the
Implementation Study.
2.39
One issue that has remained unresolved since the first NBN RFP proposal was
announced in April 2008 is that there is still no detail of the geographic
footprint of where the FTTP network will connect and where wireless and
satellite might operate. This causes uncertainty for potential investors as
well as for consumers.
2.40
The current proposal is that 90 per cent of Australian homes, schools
and businesses will have access to FTTP, while the remaining 10 per cent will
be connected via wireless or satellite. When discussing the 90 per cent/10 per
cent footprint at the Sydney public hearing, the Australian Information
Industry Association (AIIA) explained that they had spoken with Treasury
officials trying to clarify the footprint:
...we were making inquiries as to what the 90 per cent and the
10 per cent would mean. What does 10 per cent mean? It was put to AIIA that as
a rough rule of thumb it would be those parts of Australia that are populated
thinly – for example, fewer than a thousand people.[21]
2.41
The minister has been reported as stating FTTP could reach towns with
populations less than 1000 people if the necessary infrastructure is available,
or able to be readily deployed. However, in the committee's view, this
potentially adds to the level of uncertainty.
2.42
At the Hobart hearing, satellite provider Intelsat gave evidence that,
as an infrastructure provider, details of the 10 per cent footprint were a
critical issue. When asked about possible customer numbers that might be
covered by satellite, Mr David Ball answered that:
I don't know. I think the 100,000 [estimated satellite
customers] are in very remote areas which would probably fall outside the 90
per cent [FTTP] geographically. Again, it gets back to my opening question as
to how the NBN Co. defines that geographically.
...What of the 10 per cent could you serve by wireless
terrestrial means? What is the residuum that gets served by satellite beyond
that?[22]
2.43
When the committee questioned officials from the Treasury regarding the
footprint, their response was:
That is a detail that you would really have to take up with
the department of broadband. ... It is not within our competence to provide that
level of advice.[23]
2.44
Taking up Treasury's suggestion, at the Canberra public hearing the
committee questioned Mr Quigley, CEO of the NBN Co, as to the areas included
within the 10 per cent; Mr Quigley responded that:
It is scattered predominantly in regional and rural areas,
obviously, but there are also places not that far outside the metropolitan
areas that still have difficulties with coverage – ... When you look, for
example, at the satellite footprint that you might want to plan for, some of
those areas are closer in to the cities than you would otherwise expect ...
We do not have a very clear picture; we have a rough picture
at this point.[24]
2.45
Whenever the committee has pressed for specific details of the FTTP
network's coverage, the answer was invariably the same: that this will be
examined in the Implementation Study. Mr Quigley gave this detailed response to
a question he subsequently took on notice at the Canberra hearing:
Providing consolidated information on the "geographic
spread" of what is meant by the 'last 10 per cent' is extremely difficult.
Ultimately, the final details of where the 'last 10 per cent' is located will
not be known until at least the final report of the Implementation Study ...[25]
Digital divide heightened
2.46
The committee notes that the disparity of access that currently exists
between metropolitan premises and those in regional and remote communities –
the so called digital divide – will remain, despite the new policy direction
for FTTP.
2.47
The proposed minimum 12 Mbps speed for those within the 10 per cent
footprint contrasts with the 100 Mbps connection to be provided to the other
90% of the network footprint. The committee notes the digital divide will be
heightened as a result of the mandated coverage requirement for the fibre
network being revised downward.
2.48
The committee notes that as a consequence of the revised NBN policy,
some 2.2 million Australians (10 per cent of the population) will now miss out
on access to the top level broadband via the fibre network.
2.49
In addition, the government has yet to detail, for example, how schools
and educational facilities in rural and remote areas will be able to access the
same quality services that those in inner metropolitan areas will be accessing.
2.50
The committee notes that questions surrounding the issue of the FTTP
footprint and the consequential digital divide are key issues that will remain unanswered
until the final report of the Implementation Study is available, at the
earliest. This will continue to perpetuate uncertainty within the telecommunications
industry, among potential investors, suppliers of wireless and satellite
infrastructure and among Australian consumers.
Roll-in vs Roll-out
2.51
Despite concerns expressed by the committee in the two previous interim
reports that underserviced communities, particularly those in regional, rural
and remote areas, must gain access to affordable broadband before those that
receive adequate services, the government has yet to provide any assurance that
this will eventuate.
2.52
In its first Interim Report, this committee specifically called on the
government, under the terminated RFP FTTN proposal, to roll-IN the network from
those underserviced communities rather than to roll-out from urban areas that,
in comparison, are largely well serviced.
2.53
Notwithstanding the government's announcement of the regional backhaul
initiative, the committee firmly believes that this does not go far enough to
provide certainty to these communities that their needs will be prioritised.
Recommendation 1
2.54
That the Implementation Plan clearly states the government's intention
to prioritise the needs of underserviced communities, particularly those in
regional, rural and remote areas, over those with comparatively well-serviced
urban areas.
Next generation satellite and
wireless technologies
2.55
Very little detail has been provided relating to the 'next generation
wireless and satellite technology' that will provide broadband connectivity for
approximately 10 per cent of Australian homes, businesses and schools.
Wireless technology
2.56
Wireless technology, as the name implies, involves the transmission of
information using radio waves or microwaves rather than underground or overhead
wires or cables. It can be used to establish long distance backhaul, particularly
in more remote regions, or it can be used for the 'last mile' connection to the
premises or to a hand-held device. It requires an antenna on any premises wanting
to receive the transmission and numerous strategically placed base station
aerials that can relay signals across the skies.
2.57
Wireless is ideal where geographic conditions are not conducive to fixed
line cabling. For example, wireless can provide coverage over short spans of
water and across mountainous regions. This was clearly illustrated to the
committee secretariat on a site visit that covered an area within approximately
45km of the Canberra CBD. A small local service provider, YLess4U, has
installed and currently operates a successful wireless broadband network. This
currently services communities, businesses and individuals within that area, all
of whom were previously unable to access broadband due to the granite-based
mountains surrounding the capital. More details of the applications made
possible by this entrepreneurial network can be found in chapter seven.
Growth in wireless
2.58
The incidence of the wireless transmission of data has grown
exponentially over the last two years, with the rapid increase in the number of
mobile broadband connections showing no sign of slowing. Australia has been an
international leader in this trend, with the number of mobile phones in
Australia exceeding its population. In more populous developing countries, such
as India, the growth of wireless technology has been slower; however potential
for growth in those markets is enormous.
2.59
In Australia, Telstra currently has the largest footprint for mobile
telephone coverage under their 3G network and upgraded Next G wireless network.
Telstra claims that those networks cover close to 99 per cent of the
population. At the recent Telstra Annual Investor Day, Telstra's CEO Mr David
Thodey, told investors that Telstra currently had one million wireless
broadband customers and around 2.5 million fixed line broadband customers.
Mr Thodey was quoted as saying that the company expected the growth in wireless
broadband take-up to continue, and predicted that by 2015 around 60 per cent
of Telstra's broadband customers would use wireless connections.[26]
2.60
The Australian Bureau of Statistics June 2009 results on Internet usage
in Australia is detailed in chapter six. The latest figures demonstrate a
remarkable continuation of the increase in wireless broadband uptake, growing
from 1.298 million in December 2008 to 1.961 million in June 2009[27].
Subsequent to the release of these statistics, Citigroup analysts were reported
as saying:
The market ... has consistently under-estimated the wireless
broadband market over the past two years ...
Wireless broadband as a growth driver is nothing new in
Australia but the magnitude of the growth continues to surprise...[28]
2.61
The committee is concerned that the government’s requirement for FTTP technology
to underpin the NBN ignores this trend in wireless broadband uptake, impacting
the ability of the network to meet future demand.
Wireless limitations
2.62
However, the 3G network has some significant drawbacks that are
consistent with international experience. These drawbacks include the cost of
3G phones and the high cost broadband services to handheld devices. The latter
is particularly evident in more remote areas where Telstra is the only carrier
offering wireless broadband.
2.63
Additionally, although Telstra claims to cover 99 per cent of the
population, this does not equate to 99 per cent of the Australian landmass. It
is a common complaint that there is a lack of service availability in more
remote areas, along even major highways, with corresponding implications for
travellers and local remote residents alike.
2.64
A prominent industry stakeholder, AUSTAR, has been quoted as stating
that 3G networks were not suitable for NBN purposes:
[The NBN] is about a wireless data network, built and priced
to deliver data based services at affordable prices.
The 3G networks are voice networks with data as an overlay,
they don't have the capacity, the pricing structure, or the spectrum to provide
the services that are needed in the 10% areas where fibre won't reach.[29]
Wireless improvements
2.65
The industry has already moved to address some of the technical issues
with wireless technology. For example, upgraded standards have been developed,
in particular Long Term Evolution (LTE) which is based on an all-Internet
Protocol (IP) network infrastructure and uses advanced wireless technology such
as Multiple-Input and Multiple-Output (MIMO). MIMO is a form of smart antenna
technology, involving the use of multiple antennas at both the transmitter and
receiver to improve communication performance. This offers significant
increases in data throughput and link range without additional bandwidth or transmission
power.
CSIRO's cost saving solution
2.66
The committee heard evidence from the Commonwealth Scientific and
Industrial Research Organisation (CSIRO) relating to developments they have
made in wireless technology, which have the potential to be quite
ground-breaking. The solution addresses the 'last 10 per cent', and is called
the CSIRO regional access solution. The submission from CSIRO describes
the solution as follows:
In simple terms, the CSIRO regional access solution
proposes the use of existing broadcast infrastructure [base stations and aerials]
and broadcast spectrum in the new NBN. By using CSIRO technology ..., beams using
new synchronisation and co-operative networking methods will form signals over
the long distances to individual premises.[30]
2.67
The CSIRO submits that their technology is superior to 3G/4G and WiMAX
technologies, able to deliver a higher quality service with fewer base stations
at a significantly lower capital cost.[31]
CSIRO believes their development would be able to more efficiently manage
backhaul requirements in the 'last 10 per cent', providing backhaul via
point-to-point microwave radio relay. Once fully developed and patented, the
CSIRO believes that this development will not only be able to provide the NBN
with a home-grown technology solution, but will also have significant potential
in a growing international market.[32]
2.68
The cost savings estimated by the CSIRO through the utilisation of the CSIRO
regional access solution are extraordinary:
When compared with 3G/4G the capital savings are estimated to
be $12 billion; and when compared to WiMAX, the capital savings are
estimated to be $5.0 billion.[33]
2.69
CSIRO has stated that the cost of implementing their CSIRO regional access
solution would be in the order of $255 million, which would provide
backhaul services to the last 10 per cent of the Australian population.[34]
2.70
The committee acknowledges that the CSIRO is in a unique position to
provide ground-breaking, Australian-developed technology for backhaul access, advice
on the technologies, independent advice around network costs and designs and also
on applications development in the areas of health, energy management and the
delivery of government online services.
2.71
Given the obvious level of expertise, the committee asked whether the
CSIRO had been commissioned by the government to assist in the NBN rollout. The
CSIRO was careful in its responses, noting that '[W]e provide regular briefings
around our technologies.' When the committee pressed further whether they had
been asked to advise specifically in relation to the rollout in rural and
remote areas, their response was:
It is nice to be asked, and we hope we are asked, but
sometimes we are not.[35]
2.72
The committee is concerned that the CSIRO was not consulted in the
formulation of the revised NBN policy, and that insufficient consideration was
given to emerging technology prior to the 7 April announcement by the government,
which included the requirement for FTTP to underpin the NBN.
2.73
The advantage of mobile connectivity via lap tops or mobile hand held
devices to an increasingly mobile workforce is obvious. This in turn is driving
demand for wireless connectivity. In fact, there have been questions raised as
to whether the 90 per cent footprint FTTP should be more flexible, given the
move by service providers and application developers to cater for this growing
market segment. AUSTAR made the following comment in their submission:
...it does not make any sense to limit the building of a
wireless network to only 10 per cent of the population, particularly given the
benefits of portability and interoperability inherent in the wireless product.
... AUSTAR believes that the rapid deployment of a WiMAX wireless broadband
network using 2.3Ghz spectrum provides a excellent opportunity for NBN Co to
provide immediate, affordable, high speed broadband services to many
Australians.[36]
Spectrum issue
2.74
If wireless broadband is to be deployed, the government will need to
ensure that sufficient spectrum is reserved at appropriate frequencies and that
this allocation can continue to meet the demand requirement caused by the
rapidly growing uptake of wireless broadband.
2.75
The imminent digital switchover of analogue television transmission to
digital TV will result in the freeing up of spectrum previously used by
analogue TV services. There will no doubt be strong competition for the
purchase of licenses for this spectrum, with industry groups lobbying in the
media to publicly stake their claim. There are calls on the government to
ensure that at least a portion of this freed spectrum is reserved for the
specific purpose of facilitating wireless broadband. The CSIRO is a strong
advocate of this view to enable its wireless access solution:
By utilising the digital dividend of reusing the broadcasting
towers and spectrum (UHF and VHF) currently allocated to analogue TV... the CSIRO
regional access solution can deliver broadband services at 100 Mbps to
sparsely populated communities at significantly lower costs than WiMAX or 4G. ...
To deploy the CSIRO access solution, it will be necessary for
ACMA to re-allocate at least some of the spectrum currently allocated to
analogue TV. ...It is recommended that at least 35 MHz in the 400-800MHz
frequency range is assigned for fixed wireless access to rural broadband.[37]
2.76
It is unclear whether this issue is receiving an appropriate level of
consideration under the government’s Implementation Study.
2.77
AUSTAR has also highlighted that the government needs to ensure that
adequate spectrum for the wireless solution is available. AUSTAR has spectrum
that it believes would be suitable for the wireless broadband network:
...AUSTAR invested A$183 million in 2000 to obtain spectrum
licenses covering 98Mhz of contiguous spectrum in the 2.3Ghz band and ...
obtained 65Mhz in the 3.4-5Ghz band...[38]
2.78
Further, their submissions states that:
AUSTAR has made clear to the Government and to third parties
that it is willing to enter into commercial arrangements for the sale of our
spectrum licenses to facilitate the rollout of wireless broadband services.[39]
2.79
It is apparent that spectrum is as important in the facilitation of the
wireless network as the fibre is to the fixed line fibre network. However, the
issue of spectrum allocation has not been clearly addressed by the government
in relation to the requirements for the NBN. This will need to be a priority
discussion within the Implementation Study.
Satellite technology
2.80
Satellite technology is, in reality, a subset of wireless technology, which
enables global transmission of data via satellite. However, satellite and
terrestrial wireless technologies have different benefits and disadvantages.
2.81
There are three main types of satellite systems that are generally
categorised by the height of their orbit: low-earth orbit (LEO) at around
2,000km altitude; medium-earth orbit (MEO) at around 9,000 km; and
geosynchronous orbit (GEO) at 40,000 km.
2.82
Geosynchronous satellites are most common. They are seen as stationary
from the earth and have large coverage areas and consequently fewer satellites
are required and can provide a wide or even global coverage. The disadvantages
of the GEO systems are that it takes a great deal of power for data to reach
the satellite and there is the increased delay or latency issue with the
greater distances involved. Latency issues have come to the fore due to the
increased use of satellite for video, interactive games and Voice over Internet
Protocol communications. GEO satellites are also more costly to launch and have
higher maintenance costs, given the higher orbit from which they operate.
2.83
The committee heard evidence and received written submissions from two
satellite providers, Intelsat and O3b networks. Intelsat claims to be the
largest fixed satellite services provider globally, with 50 satellites covering
99 per cent of the world's populated regions. In its submission, Intelsat
states that:
...the direct-to-consumer element of the satellite NBN will
require two multi-spot Ka-band satellite payloads to ... deliver service into
multiple high power spotbeams providing contiguous coverage across the desired
service area.[40]
Footprint uncertainty concerns
2.84
The issue discussed above relating to the lack of detail of the
geographical footprint of the 10 per cent was an issue for both satellite
providers. Without this certainty, providers cannot commence planning what type
and how many satellites might be required to meet the needs of the 10 per cent.
Intelsat noted that the 'Implementation Study will need to consider a wide
range of issues relating to satellite delivery direct-to-consumers.'[41]
This concern was reiterated when Intelsat's Regional Vice-President for
Asia-Pacific gave evidence at the Hobart hearing:
One of the first activities that has to be undertaken by the
department or by NBN Co is truly defining where the 90 per cent is
geographically and where the 10 per cent is in terms of the 100 megabit and the
12 megabit definitions that have been provided. ... Is it a traditional satellite
that is in orbit today that can provide services to customers, or is there a
much larger number of customers that need to be addressed that lead you to
putting up a KA band satellite which would be dedicated to NBN?[42]
Eliminating the latency issue
2.85
Mr Greg Wyler from O3b Networks gave evidence at the Melbourne public
hearing and explained very clearly the operation of satellite and the issues
that his satellite system could overcome. O3b stands for the 'Other 3 billion',
a reference to the people in the most remote and least populous regions of the
world who are currently not able to access the Internet, let alone broadband
services.
2.86
O3b is designing a satellite infrastructure system that will see eight MEO
satellites launched in 2011 that will orbit at around 8,062 km above the equatorial
belt, with coverage of plus or minus 45 degrees north and south of the equator.
Mr Wyler stated that O3b's network would have beams that would be
'steerable', whereby the footprint of the satellite can be altered to meet
community and capacity requirements. Mr Wyler outlined what he believed to be the
advantages of O3b's satellite system, as follows:
Geosatellites have been very expensive – about $250 to $300
million to produce. Our satellites are ... about $22 million to produce. We put
them in orbit at I think about 8,062.7 kilometres, which is 4.6 times closer to
the earth than a geosatellite. The satellites being a lot closer to the earth
means that they use a lot less power to bring just as much capacity to the
earth or to the customer. On top of that we have allocated 4.3 gigahertz of
capacity, which is much more than any of the geosatellites.[43]
2.87
The MEO satellite system deployed by O3b would immediately address the
latency issue that currently inhibits the quality of services. Mr Wyler suggested
that an additional four satellites would most likely be needed to provide
coverage for the Australian landmass, including Tasmania, at a cost of $150
million to $200 million for those four satellites. This is less than the cost
of just one GEO satellite.
Satellite limitations
2.88
Both the GEO and MEO satellite systems provide a highly reliable
connectivity, 'even in comparison to fibre and microwave'.[44]
However the one issue that neither GEO nor MEO satellites can eliminate is the
susceptibility of satellites to adverse climatic conditions. Both O3b and
Intelsat suggested that Australian satellites should utilise the Ka band
frequency, which unfortunately is more affected by rain than other frequencies.
2.89
Mr Wyler explained that the susceptibility of Ka band satellites to
weather can be overcome by strategic system design that would manage this issue
and minimise the impact of adverse weather patterns. This would be necessary in
Australia, given that the O3b satellites would orbit over the tropical
monsoonal areas. However, as pointed out by Mr Wyler:
Nothing is foolproof. There can be moments of outages in any
[network] topology...
Satellites can certainly be designed to have very limited
sensitivity to weather conditions.[45]
Oversubscription and contested
networks
2.90
Another disadvantage with both wireless and satellite technologies is
that the services received are 'contested'. The speed that the network offers
is a maximum speed, and is shared among the number of users in that particular
network. This only becomes an issue in more populous areas, where more people
are likely to be using broadband services simultaneously.
2.91
Both Mr Wyler and Mr Ball warned that for this reason, any operator of a
satellite network must take into consideration the oversubscription limitations
under which their network can effectively operate. Mr Wyler gave a clear
example of the effect of oversubscription:
Oversubscription is a key factor. ...
If you provide 12 megabits piped into a town and then put up
a WiMAX tower and then offer everybody 12 megabits to their home and you have
100 customers, then you have 100 times oversubscription. You have promised 100
people 12 megabits, but the reality is there are only 12 megabits ... Unless
you watch the oversubscription, or at least take note of it, it is possible to
find that the quality you are hoping for is not really achieved.[46]
2.92
However, as Mr Ball explained, this issue also applies to ADSL
technology and to any wireless technology, and needs to be monitored and
managed:
[Oversubscription] is one of the challenges you have with any
wireless technology as you start to load it up, and indeed any ADSL type
technology today. As you start to oversubscribe point of presence, you end up
having to provide additional capacity.[47]
2.93
Mr Wyler concurred with this view:
There will definitely be oversubscription. There is
oversubscription in every network; it is inherent. But you do not want too
much.[48]
2.94
Adding satellite capacity in order to address oversubscription in a
region is a long term, expensive project. Both O3b and Intelsat stated that it is
generally a three year process from the planning phase to launch of a satellite
service. Mr Wyler stated that consequently, it is critical to gain a clear
understanding of the population densities and the capacity requirements of each
area, to ensure that the oversubscription issue is managed to achieve and
maintain quality service.[49]
2.95
The committee believes that, given the vast geographic expanse of the
Australian continent, its varied climatic conditions, and its diverse topography,
the technologies of both wireless and satellite should be considered as
complementary to the FTTP network.
2.96
In addition the committee believes that there is scope for greater
flexibility in the FTTP footprint, and that the technology/ies best suited to a
particular location should be selected for deployment in that location, thus
ensuring optimal quality broadband services to all Australians.
2.97
The committee also calls on the government to the release details of the
90 per cent/10 per cent footprint as early as possible to enable
wireless and satellite providers optimal planning capabilities, which will in
turn provide for informed estimates of the cost of deploying the 90/10 network.
OPEL?
2.98
In discussion of wireless and satellite technology options, the issue of
the Coalition Government's cancelled OPEL proposal was raised a number of
times, generally in reference to the fact that, if allowed to proceed, the OPEL
solution would have almost been fully deployed at the time of reporting.
2.99
The OPEL network was a joint venture partnership between Elders and
Optus that, among other services, would have utilised wireless technology to provide
improved backhaul access. Specifically targeting underserviced areas through a
subsidy program, the OPEL proposal would have delivered ADSL2+ broadband
services for 1.5 million premises in regional areas by upgrading an additional
312 exchanges.
2.100
The OPEL proposal was never initiated, so the doubt cast by the
government on the viability of OPEL's solution cannot be tested.
2.101
Although the current NBN is promising higher speeds to regional areas
than OPEL offered, the committee again states its concern that the
underserviced areas still have no guarantee that they will receive priority
attention in the roll-out implementation plan.
2.102
Despite the government's claims of providing better broadband services
to regional and remote Australians than the OPEL initiative, the committee
highlights that under OPEL, these underserviced areas would now be accessing
broadband at ADSL2+ speeds. This would have been a vast improvement to the
government's record of inaction and the status quo – which could persevere for
the full length of the NBN rollout.
2.103
The committee notes that due to the lack of available cost benefit
analysis data, it is difficult to quantify whether the rural and regional
component of the $43 billion NBN commitment is a positive value
proposition compared to the $1 billion in public funding required by the OPEL
initiative.
Committee view
2.104
The committee remains concerned whether the 90/10 footprint will meet
the demand profile for broadband services now and into the future and is
particularly concerned that the 90/10 footprint has not been clarified for the
Tasmanian roll-out.
2.105
Despite the commencement of roll-out, Tasmanians are still in the dark
as to which towns the fibre will touch and which it will by-pass.
2.106
The committee acknowledges the multiple complexities facing the
government as it makes decisions around the architecture that will provide the
optimal solution, both for the NBN Co and for all Australians.
2.107
The architecture is one of the key components in determining the cost to
the tax payer of this network, as discussed in chapter five. Hence an early
decision on the network architecture will enable a rigorous cost-benefit
analysis to be undertaken.
2.108
The committee notes the significant cost and energy efficiencies to be
gained by deploying GPON architecture as compared to a P2P network; however, the
committee also notes that P2P architecture provides greater scope for service
differentiation.
2.109
Importantly the committee highlights the continuing rapid growth in the
proportion of wireless broadband connections and questions whether the 90/10 percentage
for FTTP and wireless/satellite connections should be more flexible to leverage
this increase.
2.110
Evidence brought before this committee has stated that a wireless
broadband network could be deployed relatively quickly, particularly if optimal
use is made of existing base stations. In this manner, wireless could be seen
as a 'first step to address the long standing needs of underserved
communities.'[50]
In the longer term, wireless and fixed-line fibre will be complementary
components within the NBN. The interrelated issue of the allocation of
sufficient spectrum to enable the wireless deployment requires urgent attention
and resolution.
2.111
The committee is particularly concerned that substantial savings, in the
magnitude of billions of dollars, are a very real possibility when the CSIRO
solution for backhaul re compared to the alternatives of WiMAX and 3G/4G
deployments. The committee is concerned that the government seems totally
unaware of these savings, following the recent commencement of a major WiMAX
rollout in South Australia as part of the Regional Backhaul Blackspots
initiative.
2.112
The committee urges the government to consult with CSIRO prior to the
awarding of any new regional backhaul contracts to determine the feasibility
and possible development timeframes involved in this cost-saving and superior
backhaul solution.
2.113
The committee also urges the government to consider the alternative
options for satellite deployment featured in this chapter.
Navigation: Previous Page | Contents | Next Page