-
Website
http://www.inquisitr.com/ -
Original page
http://www.inquisitr.com/21346/australia-commits-au43-billion-to-broadband-network-that-will-be-redundant-before-its-finished/ -
Subscribe
All Comments -
Community
-
-
Top Commenters
-
Popular Threads
Like the Cable will hit a high speed this year but its based on better tech not new Cables. Surely fiber speed will get quicker over the near term future.
Just a thought and could be completely wrong.
Yep, so it's good if you live in the capital cities, but if you live anywhere else... too bad?
Plus, the more competition Telstra has the better.
VDSL holds more promise, but that 500Mbps is only possible within about 500 metres of the DSLAM/exchange/enabled-RIM. VDSL breaks down to essentially ADSL speeds very quickly as you move further away.
Fiber has the capacity for very long runs. The residential stuff they use for FTTH installations is, however, a plastic-based very cheap to install derivative. This is only good for reasonably short runs also because of the high attenuation. Coupled with the up to 128x half-duplex multiplexing that is used for residential installations, you start to run into significant limitations for providing any more than the 100Mbps proposed bandwidth. If you wanted to provide more, it would require a physical upgrade of optics at both the customer and the exchange end of each single piece of fiber (actually per Point-to-Point connection... multiplexing only works with optical frequencies, even if you are multiplexing 128 channels across a single piece of fiber, you need 128x optic couplers at each end to read/write).
Not that 100Mbps is to be sneezed at. You have to remember that it is possible to deliver this symmetrically (the VDSL mentioned above is asynchronous meaning the upstream bandwidth is significantly lower than the downstream). A lot of interesting things can be done with 100Mbps of symmetric bandwidth, especially with cloud-computing and mesh-style delivery of content.
Even without interesting use of everyone being connected with 100Mbps of symmetric bandwidth, 100Mbps is still good for 2-3x 1080p streams, phone and high-speed Internet. This opens a world of possibilities (of course, this would all have to be delivered relatively close to each home because providing backbone infrastructure to deal with 10 million homes requiring a 100Mbps connection to the USA would be significant - 1Pbs (or about 500x bigger than any inter-country backbone in existence today).
Sorry but very very wrong. The fibre used in residential installations is not plastc based and is not cheap.
"This is only good for reasonably short runs also because of the high attenuation"
Again very wrong. The technology used in FTTH delivers 20+ kilometre reach.
"Coupled with the up to 128x half-duplex multiplexing that is used for residential installations"
Again very wrong. It is not half duplex multiplexing. It is a time and wave division multiplexing. Wave division on the direction, time division on the access. Most deployment today are based on 32 and sometime 64 way multiplexing
"you start to run into significant limitations for providing any more than the 100Mbps proposed bandwidth"
Again very wrong. FTTH solutions today are delivery greater than 100Mbps using PON technology.
"If you wanted to provide more, it would require a physical upgrade of optics at both the customer and the exchange end of each single piece of fiber"
Again very wrong. The upgrade path of 1Gbps or 2.5Gbps to 10Gbps PON is an inplace upgrade with a mix of both 1/2.5Gbps and 10Gbps on the same pon at the same time.
"VDSL holds more promise, but that 500Mbps is only possible within about 500 metres of the DSLAM/exchange/enabled-RIM"
VDSL does not hold more promise. The 500Mbps which was only demonstrated the other date by Ericsson as opposed to the 36m deployed FTTH connections around the world. VDSL2 does not deliver 500Mbps at 500meters, its 300Mbps. At 1000m is getting down to ADSL2+ speeds.
If you are building FTTN using VDSL2 and want 100Mbps to every home, you have to place the nodes every 300m. In that case you are already passing every single home with fibre to feed the nodes.
32 to 128-way is possible with DWDM and TDMA and is used for FTTP (in fact 32,768-way splits are theoretically possible with EPON).
From my understanding of most FTTP installations, a single strand of fiber is installed to each CPE (not a pair), which is what I was getting at with half-duplex. There is a single piece of fiber used for both directions, instead of one for up and one for down (which is how most fiber is installed). This also makes the provision of symmetric speeds more complicated (and potentially less likely) as it requires the OLT to allocate upstream bandwidth to the ONT to prevent collisions. If a network wants to get more bang for their fiber buck, they can allocate more bandwidth downstream than upstream and run more multiplexed connections.
I wasn't aware that PONs only had a single optical connection at the ISP end of the fiber (which obviously makes upgrading simpler) however, it would still take an upgrade of all the optics at the customer (and still, the provider end) to increase capacity past whatever optics are installed to begin with (which seems like it will be 100Mbps)
In regards to VDSL, I was referring to it holding more promise than cable-based technology because it is P2P between the end-user and the DSLAM. Not that it holds more promise than FTTP which is dramatically better for obvious reasons.
Copper is NOT a holy grail. Fibre is, and always has been. You must have fibre within 100 metres of the premises to even get 100 megabits over copper. So what do you need for 500mbs? As the linked article states, VDSL is only useful for the last mile (more like last 100 metres).
Once you have fibre in the ground, you can upgrade all the switching gear to support 1Gbs speeds. Copper will hit a speed ceiling long before fibre will.
Plus, as the Aust. Govt consults about this over the next 10 months, they may realise they need to build with 1 Gbs in mind.
busy training firefighters maybe?
Also comparing a country like Australia to a country like Japan, is madness. The size and density are not comparable.
If I was a government organisation, I would put high-speed infrastructure in places like this first (reasonably high density with few options) and save the well-serviced (for competitive ADSL2) inner-city areas for last.
You do realise that they are rolling out fiber?
What do you think they are using in other countries to achieve speed of over 100Mbit/s at home: fiber...
The example with Japan is one particular example, they are rolling out fiber too.
Fiber already supports 10Gbit/s today, how much faster will it support in 2018 ?
Upgrading the network from 100Mbit/s to much higher speed will only be a matter to upgrading the nodes ; just like you could upgrade from ADSL to ADSL2+...
FTTH is the most logical replacement of the copper network...
Better sooner than later.
Japan is busy converting from DSL to FTTH.
As of 2008, they have more broadband subscribers on FTTH (45%) than DSL (42%).
See:
http://www.soumu.go.jp/menu_news/s-news/2008/08...
Guy must own telstra shares or something. Don't know what he is worried about as one would assume Telstra will bite the bullet and help build all this at some point.
Perhaps the Inquisitr will assign the story to someone who actually knows what they are talking about when they next broach the subject of the NBN.
> The problem with this huge investment is that the network will be redundant before it is completed.
The maximum bandwidth of an optical fibre connection is, in the context of a home or small business connection, essentially unlimited. They may start at 100mbps, but change the equipment of each end of the fibre and it can go much, much, much faster than that.
> Australia’s largest telco Telstra is already rolling out a 100mbps network in capital cities, beginning with Melbourne this year.
Sure, to a much smaller set of homes than the proposed FTTP network. Extend that cable network to cover 90% of homes in the country and suddenly it's going to get a lot more expensive. Not to mention cable is a shared bandwidth medium, is not synchronous, you're not solving competition issues like the FTTP proposal is, etc.
> But those are before we get to the holy grail: telephone lines. 100mbps over copper with DSL is already a possibility. 500mbps VDSL2 over copper (that is, the telephone connection into a home) is already being tested
If you live within 500m of an exchange. And your telephone line is in good condition (unlikely given the lack of maintenance in recent times). And you don't care don't about the other aspects of the FTTP build like monopoly (Telstra) breaking, etc.
> Imagine the competing technlogies and services on offer by 2018?
Ahh I see... so we should wait until 2018, because by then the new technology that will make today's technology redundant will be around? But what about the technology that will make 2018's technology redundant by 2028? Better wait for that too. And 2038's, yada yada yada.
> South Korea is rolling out 1gbps over 3 years today; with some advances in technology, 1gbps or even faster may be the norm in Australia when the NBN finally finishes its 100mbps rollout 9 years from now.
OK, now I'm sure you either did no research, or have an interest in talking down this FTTP plan.
How do you think South Korea is rolling out 1 Gbps? That's right... FTTP! They're just using gear on each end of the fibre that can handle higher speeds than what is being /initially/ proposed in Oz.
This is one of the best (the best?) telecommunication policies in Australian history. Fibre is /scalable/ - this network will serve our needs for a long time into the future. It will create jobs in the short-to-medium term. It will drive economic growth, improvements in healthcare, education, etc, in the medium-to-long term. And best of all, it will prevent Telstra from continuing to hurt competition and innovation in the telecomms industry in Australia by completely removing them from the infrastructure equation. Good stuff.
when copper networks first started being used for data, we were getting 300bps, now we have 24Mbps ADSL 2+ commonplace. in 8 years, we may see gigabit speeds in fibre, using whatever technology the ISPs decide to use over the fibre network.
this article is short sighted and shows a clear misunderstanding of the technologies involved.
nation building it most certainly is.
Of course, I'm yet to see a way to use all this bandwidth. Is 1080p high definition TV used by anyone right now (30Mbps)? Is there even anything else being created to replace/extend 1080p TV? Is there any other consumer application for > 100Mbps? Maybe live content streaming for games (of course, what's the point, wouldn't it be easier to just install 100GB, 1TB, 10TB of textures locally)? How long will take consumers to even upgrade past 100Mbps home networks (or worse, 802.11g running at < 20Mbps)?
The only thing controlling the 100mbps speed is the electronics at each end, which can be easily changed to vary the speeds.
VDSL2 at 500Mbps is great, but what about people with aging phone lines and 2+kms from the exchanges?
Are you saying "Dont go ahead with the NBN?"
and referring to south koreas 1gb connection plan have you even considered how much larger australia is geographically to even compare the two is idiotic
then also the fact that fibre is currently as good as it gets as a medium (cable), its biggest problem is the fact that the hardware it connects to is way to slow
if you can get 100mbps from copper wires, whats the potential for optical
worst
article
ever