Slashdot Mirror
Aussie Telcos Consider 3G For Last Mile
Mattygfunk writes "Whirlpool reports Aussie telco R&D labs are scrambling to transform 3G mobile networks into last mile solutions rivaling the best wired broadband networks, as telcos come to grips with lack of consumer interest in 3G mobile services and a likelihood of no payback on their multi-billion dollar investments in the spectrum."
wow, finally an affordable application. in europe, 3G almost ruined a major player in the telco market, and severely damaged the financial situation of the German Telekom, after the govt made some 60 billion EUR.
(Oh, as a small remark.. German Telekom's shares are in public hands.. )
The european 3G band is 1.9-2.2GHz.
Infra red is 150000GHz to 400000GHz.
AT 3G frequencys the signals are considerably attenuated by passing through objects like walls and people, typicaly 30dB per wall. Trees with wet twigs and leaves that are about a quarter of a wavelength (4cm) create significant rf shadows.
I'm just a radio ham who experiments with such stuff but my back of an envelpe calculations suggest they are going to need lots more base stations (since higher power is unlikly) to get the same coverage as GSM, paticularly inside buildings.
At AU$0.20/Mb for anything over the base plan, data costs are what's keeping the internet expensive here. The current pricepoint is AU$80-90/month with a 3Gb cap. It doesn't matter what the delivery medium is used to deliver the data, that cost sets a lower limit for any pricing scheme.
Cable is severely restricted to the highest areas of population density in the south-east of the country, ADSL is available in a high percentage of exchanges(both metropolitan and regional), and satellite is used for those areas where copper doesn't reach and can afford it($$$$$).
A 3G infrastructure depends on a good underlying data network - which basically describes ADSL(plus ISDN, ATM, etc).
As far as I can see, the areas which have that infrastructure, but where 3G is the only solution are extremely limited. Add that to the very limited market of high-speed-mobile-data users, and you have a solution which has an extremely small number of users for a high development and setup/intrastructure cost.
I doubt it will be a competitive or affrodable "last mile" solution for the forseeable future, outside of a few PR "success stories".
This is correct. Higher power is used, but I believe one estimate is that to get comparable coverage to a GSM network requires up to four times as many base stations.
The situation is further complicated by the fact that as a result of the code-based multiple access technology the effective size of a cell varies with the number of users connected, making the design of networks with good coverage non-trivial.
I used to work for Ericsson as a RF Engineer for UMTS networks (though these opinions are, of course, my own) and, although I would quite like to be proven wrong, I get the strong impression that 3G will never deliver. The "2Mb/s" bandwidth is a theoretical maximum for a terminal right next to an otherwise unused base station. 365 Kb/s (asymmetric, with 64 on the uplink) is a much more realistic figure. Providing even this with good coverage requires an enormous number of base stations.
Even with groovy multimedia apps, there is a limit to how much people will be willing to pay for this technology. IMHO (IANAA - I am not an accountant), this is going to be a hell of a lot less than the many billions which have already been spent on licenses and networks.
Cue Chopins "Funeral March".
... I don't think it'll be Telstra doing it. By the looks of it, the Australian Federal Government is going to sell Telstra whether or not it meets its 'bush coverage and standards' requirements (they'll fudge the figures to make it look like it has passed). A privatised Telstra will have no interest in the bush.
Over the next 12 months, Australia (or at least the main cities) is going to see many 802.11b ISPs set up shop. With Australia's low population density (making it expensive to have cables running to every home) and Telstra's reluctance to provide even semi-decent infrastructure, I think this sort of thing will be the future of Australian telecommunications.
In the end, this should be a good thing, since competition will be greater. Compared to a cable-based (telephone, fibre optic, etc.) market, there are far fewer barriers to enter a wireless market (the hardest part is getting a spectrum license). Scores of smaller companies will enter, and we will no longer be as subject to a predatory, uncompetitive monopolist (read: Telstra) as we are now. Only time will tell if that situation will last, though.
I'm pinning my hopes on ultrawide broadband (UWB) communications. Bob Cringely wrote earlier this year that it has the potential to be simultaneously ultra-fast, secure and frequency-neutral (i.e. since it uses the whole spectrum, a virtually unlimited number of UWB signals can be running concurrently). The only problem is the range (about 1km, I believe), but 802.11b has this problem, too.
OLPC Australia
not a teleco person, but let me answer still.
Afaik WCDMA cell can scale from 512 users at 8k/s to 2 users at 2mb/s. Sounds like too little? Well, the cell size for WCDMA is significantly smaller than on gsm, and as the voice calls will be mostly routed through the gsm network, the improvement should be pretty visible. But definetly no warezing channel.
What I'd like to know however, is the roundtrip time over different 3G technoligities. Even GPRS has adequate bandwith for browsing/mail/messaging, but the 1 second roundtrip time with The TCP 3-way handshake really kills the experience.
WCDMA is the non-qualcomm version of 3G, as qualcomm got too greedy with it's CDMA tecnology patents. Yet another case in to the "patent system needs to fixed" bucket. Someone else should write about how much users their CDMA2000 can handle. (The UMTS version, not the 1x version).
3G is just one step that provides extra wireless bandwidth. Next step would be roaming from high-speed WLAN networks to larger coverage 3G networks, and finally to The everywhere covering satellite network. The obstacles are mostly commercial, as everyone want's start billing everyone else for the roaming.
signatures pending - ansa@kos.to - (dont mail there)
bandwidth per cell is 388,800kbps
resulting in connection speeds of
2000 kbps for 194 subs
384 kbps for 1013 subs
144 kbps for 2700 subs
64 kbps per 6075 subs
Note cell size is > 1Km but I don't have a figure for max cell size.
Source was from a industry report from an investment house.
384Kbps would probably be the most cost efficient and marketable solution giving a similar user experience as ADSL. However cost per Kbs will be hard to get as low as the monthly connection charges and unlimited connection time we see in most ADSL contracts these days.
Still, any to get money back into telco's cant be bad for the industry as a whole.
If you get modded down for a first post... What do you get for a last post?
A little company by the name of ArrayComm is currently testing a technology they call I-Burst here in Sydney. Basically, it (should) offer pervasive, roaming wireless internet access across every major city in the country, at 1Mbps per user.
What does this mean? Well, it means that if you've got I-Burst capable NIC's, you'll have a 1Mbps Internet connection at any point across all our major cities (as long as you're not moving too fast) on your laptop, PDA or even your phone (assuming there is phones to take advantage of the tech - doubtful perhaps).
This is beyond what any of the 3G's (W-CDMA, CDMA2000 etc) can offer. The only reason Australian telco's are pimping 3G is because they all spent small fortunes purchasing spectrum to run it on. Considering that these 3G solutions are a long way off, and that we should have a working commercial I-Burst service here in Oz sometime next year, 3G may just die a very quick death.
We conducted an in-depth breakdown of the technology behind I-Burst, including the special directional antennas that make it possible. You can check it out here if you're interested.
Janie took my gun...
When the Europeans were deciding what their 3G telco standards would be, there was a division between Siemens (who wanted to use a Time Division Duplexing System called TD-CDMA) and Nokia and Ericsson (who wanted to use an Frequency Division Duplexing system called W-CDMA). Rather than choosing a winner, the European Telecommunications Standards Institude (ETSI) adopted both standards, and technically UMTS supports both, although W-CDMA has got all the publicity. Spectrum bands were allocated for both, paired (for W-CDMA) and unpaired (for TD-CDMA). In Europe, telcos are required to use these standards, but there certainly have been allocations of unpaired spectrum. Australia decided to auction the same spectrum bands that were auctioned in Europe, but put no restrictions on the technology they were used for. 3G licences were sold in Australia for not much money, as the number of licences and number of players were the same. Arraycom bought some unpaired spectrum for not much money, and always said that it would use the spectrum for its i-Burst technology. This is now happening. (The article is false when it said that Australia uniquely allocated unpaired spectrum. Most of the world did. Australia merely put uncommonly few restrictions on its use. This was not unique either, as other countries (eg New Zealand) did the same.
I don't think i-Bust is technically '3G', but it does use '3G' spectrum.
Michael.