Slashdot Mirror
Australian Researchers Push Near-Broadband IP Over VHF
Curmudgeon Rick writes "A research group at the Australian National University is getting symmetrical 250K bps at 20km, using "empty" 7MHz-wide broadcast TV allocations in the 45MHz band. Story here, project homepage here. Aim is to put some bandwidth out beyond the reach of the wires, where users are few and far between."
Mesh networking would be a better idea than all of this. More bandwidth, more parallelism, less power.
It doesn't sound as if they are really ready to talk about frequency coordination with other users. I hope they don't go about asking for spectrum for anything but experimentation this early in their project.
Bruce
Bruce Perens.
ANU... go to bottom of page
Conquering that 'last mile'
Pioneering work by physicists and engineers at ANU to build a cheap, simple and robust wireless communication system may soon see regional Australia getting a workable connection to the Internet. The system is called BushLAN, and it's all about bridging that 'last mile'.
Regional Australia has never had adequate access to the Internet. It's either not available, too expensive or unreliable. A major part of the problem is the 'last mile' of access. This 'last mile' is the connection between the central communications hub in a local town to individual residences and businesses. Unfortunately, the 'last mile' is usually much more than just a mile. In rural areas such as Cowra, for example, the last mile has been measured to be anywhere from three to 100 kilometres from the town centre. In more isolated areas it can be much greater.
The cost of cabling to only a few customers over these distances is prohibitive and current wireless solutions aren't practical. Satellite connections are expensive and usually require a cable connection for a user to send information out (ie they receive downloads from a satellite but send information out via the telephone). There are ground-based wireless connections commercially available but these operate in microwave frequencies using directional antennas that require a clear line of sight to function. Given Australia's sparse population and frequently hilly terrain this would require a large number of repeater stations.
Dr Gerard Borg is a plasma physicist at the Research School of Physical Sciences and Engineering. His work with radio transmission has convinced him that the last mile could be effectively bridged using the low-VHF radio spectrum. This part of the radio spectrum has much longer wavelengths than the microwave frequencies used by other wireless systems and this allows signals to be transmitted further without the need for expensive repeaters or satellites. What's more, it doesn't depend on line of sight as the signal has the ability to go around mountains and other large obstacles in the landscape. At the moment the low VHF radio spectrum is used to transmit TV signals but with the decommissioning of some analogue TV bands in 2008 (digital TV uses higher frequency radio) there's an opportunity to switch this unused spectrum over to data connections for regional Australia.
BushLAN (Bush - Local Area Network), as the system is called, has the potential to provide remote users in regional Australia with a permanent, high-quality Internet connection (at more than 100 kb/sec) at an affordable price. However, to get BushLAN up and running, many technical and marketing aspects of this multi-faceted system have to be developed first. To achieve his goal, Dr Borg has enlisted the assistance of a wide range of students from the Faculty of Engineering and Information Technology who have taken on the various jobs associated with the system as part of their Honours, Masters or Doctoral projects.
"The practical nature of BushLAN and its relevance to regional Australia really attracts the students," says Dr Borg. "Once they're involved, they become highly motivated about what we're trying to achieve. Quite often they finish the formal part of their work for their thesis, but then they stay on working on the project through the Christmas vacation."
The next step for BushLAN is to set up local trials to test transmissions, and then work with interested Internet service providers to see how BushLAN can be integrated into existing information systems. The hope is that with BushLAN as part of the system, the 'final mile' will no longer be an unbeatable hurdle.
Science Reporter is brought to you by the National Institute of Bioscience, the National Institute of Engineering and Information Sciences, the National Institute for the Environment, the National Institute of Health and Human Sciences and the National Institute of Physical Sciences. Written by David Salt.
For more information on any of the stories presented here please visit http://ni.anu.edu.au/
Quidquid latine dictum sit altum viditur
In Australia the ACCC defines BroadBand as 200kps and over.
er, that would be no.
The channel-width is 7MHz, the frequency is around 45MHz.
Visit CryptoGnome in his home.
Narrowband: less than 2Mbps (POTS, ISDN, etc)
Broadband: greater than or = 2Mbps.
Most DSL lines work at under 1Mbps (home users don't need more and it's damned expensive if telcos offer it at all), but in full flight, it can reach around 8Mbps, so it's technically broadband.
One man's *near* broadband is another man's 2B+D.
A T1 is 1.544Mbps. Not sure where you got your definition of a T1 from, but i'd suggest you check it.
Generally the problem using VHF is that its pretty limited to a few miles unless you are pumping some serious Watts with a gigantic antenna. Also I am pretty sure 45MHz is used for Military VHF communications(I believe the band to 30MHz to 87.975 or some such) but I am not sure about Australia and its Military freqs. I seem to remmember we could send "data" over VHF when I was in but it was hardly used(I just don't think anybody knew how) We did a bunch of teletype stuff that was pretty annoying.
Back to the topic we did hve some mobile but high powered VHF stuff but it was pretty much limited to LOS(line of sight) maybe if your lucky 30 or so miles.
HF would be more impresive because you could bounce that sucker half way accross the world but you crazy antenna configs like sloping V's and such. Then you would have to worry about weather, TOD, and the infamous sunspots.
I am always surprised that satalite internet never really took off in that more people could access the same bird regardless of remote thier location and steup is pretty minimal while the range is outstanding.
But really sucks is when you have to carry them...
It's all Politics
As a transmission engineer, you should already know that, in the US, that part of the spectrum (45MHz +/- 3.5MHz) is already allocated and HEAVILY used by all kinds of two-way land mobile radio systems. In fact, the band is sliced up so that even various Federal agencies (including the military) are using it. TV starts at 54MHz, and those frequencies (TV channels 2-59) will be used for HDTV.
Go read 47CFR Part 2...
FAQ...etc.
CC.
TaijiQuan (Huang, 5 loosenings)
Digital TV, including HDTV spectrum, has already been reserved slightly higher up the band. The band Dr Borg is using will be vacated by current TV stations. Since HDTV and SDTV will be required to be running side by side with analog TV before the 2008 deadline, there is no point in using anaolg TV spectrum for HDTV channels.
Aside: the Digital TV in oz seems to be running close to maximum compression. I had a fairly decent antenna setup, so I find the mpeg artefacts in digital tv more annoying then what I had.
Yes they do. I am a licensed amateur radio operator (with moorse code, a nifty skill to have, though now kinda useless lol). It is called packet radio. I myself haven't gotten into it, but it is a different protocol which you can tunnel others through (tcp/ip). Some are internet repeaters, basically hooking up their cable modem to their ham radio. Problem is, that only one person can transmit at a time, collisions aren't too common though.
If you are about to mod me down, keep in mind that this post was most likely sarcastic.
This might fly in Australia, but probably not in the US or other large nations. The radio spectrum is a limited resource and as such, a highly competitive one. Amateur radio operators (myself included) are constantly trying to defend our allocated bands here in the US against commercial entities who would like to have it for their own usage. I don't see a system that uses this much bandwidth being practical for US usage.
17W of power to get 40km?
I use/test/setup equipment that goes 50km at 0.5W of power. OK, they're using a non-optimal antenna, but the antennas really aren't that expensive. 17W would kill a bettery quick. No surfing for pron at night anymore.
200kbps is interesting, and as Bruce Perens mentions, they should have been able to get a lot more bandwidth out of the spectrum they are using. None-the-less, they could be making a robustness/raw data rate tradeoff in the modulation scheme. I am probably just blind right now, but I can find any real details on any of the linked to.
One comment above askes about inter-city propgation, and whether this technology would/could be an answer. I think 3G and or 802.15(MAN) are better because they have been designed for intercity communications, and meet or beat the 200kbps mark, which much less power used (this is very important!)
That's not entirely true.
In the land Down Under, Telstra is the dominant telco, and it's currently 51% government owned. The current Liberal Party (think: nice Republicans)-National Party (think: farmers) coalition government really wants to flog off the rest. The problem is that Telstra provides many services to the underpopulated areas (aka "the Bush", who are generally represented by the National Party half of the Coalition) that really don't make much economic sense but make a lot of political sense. Also, it's sort of halfway decent that the outback farmers get at least a phone service. Anyway, every man and his dog knows that if Telstra gets fully privatised, *bang* there goes any semblance of service to the bush, since it is just not econmical.
To that end, the government has brought in a Service Guarantee (including Universal Service Obligations) that says (amongst other things) Telstra must provide certain minimum standards to all subscribers, and if they don't they get smacked. The government hopes that after a few years we'll all see what a good corporate citizen Telstra is and give the Libs the OK to flog off the other 51% of Telstra.
Now, one big complaint from the bush is that they get bugger all access to broadband. Even getting net access at all can be tricky for them. Satellite (if available) is very expensive. This would almost certainly not improve under a toally privatised Telstra. However, if Telstra could provide near-broadband to the bush without having to string up hundreds of miles of cable, things would again be looking promising for the privatisation thing to be on the agenda again.
Speaking from a purely Australian voter/taxpayer POV, the keyphrase is the National Party might be the junior member of the coalition but they can wield a fair amount of power over the Libs when they want to.
.When they came for the communists, I said "He's next door. Take him away. Goddam commies."
Close. it is actually the government doing the dodging. Telstra is currently partially owned by said government (they have already managed to sell half of it) and various political machinations are involved in selling the other half.
Many people outside of the major population centres are concerned that a fully private company would have little interest in supplying them services. Said company would get much greater return on their investment by solely targeting capital spenditure in the few major cities.
The only two real assets that Telstra has is the enormous physical infrastructure constructed by tax payers dollars and its current customer base. As the current government has already partially privatised Telstra, they can no longer do the only sensible thing and split the company such that the physical infrastructure remains publically owned and (partially) funded. All service providers would then stand on a more level playing field.
Analogue television channels in Australia are 7MHz wide. The channels of interest are between 45 and 75MHz. BushLAN is not necessarily tied to using a particular block of spectrum, or an entire television channel. BushLAN subdivides available spectrum into 300kHz channels. As always, there is a tradeoff between transmitter power, communications range, and the data rate.
Using two 300kHz channels [for a symmetric full-duplex connection] low power, relatively short range links with a raw data rate of 115.2kbps have already been created.
As to propagation. VHF achieves beyond line of sight range whereas microwave links are limited to LOS. Long distance propagation is largely due to diffraction over hill tops. Atmospheric attenuation is much smaller at VHF (wavelength is roughly 6 metres) than it is at microwave frequencies (wavelength: ~10cm). This allows greater reliability during adverse weather conditions.
Actually fiber would generally count as baseband. It depends on if you count DWDM as part of the 'broadband'. The highest speed base fiber systems (OC-192 or 10GigE) are baseband in that a laser flashes full on or off to indicate the data on the line. (Usually flashes 'on' to indicate a change from the previous state, or to indicate a maximal clock period has passed)
DWDM is really multiple baseband signals carried on the same fiber, in the same way you would carry many ethernet signals on a bundle of 25 pair cable. (in this case color)
To be truely broadband, the laser would have to encode information into a band/color shift.
If you don't have line of sight, I wouldn't hold your breath at 2.4 gigs. I wouldn't bother wasting time trying with laser.
Both technologies are inherently line of sight, 2.4 gz will suffer severe attenuation by anything in its path. It is possible to cause signals at vhf/uhf to bend over hill brows, but the signal will be attenuated severely due to scattering.
20 Km distance at 2.4 gig is going to need some fairly high gain antenna at both ends, even with line of sight, the figures for free space attenuation at this frequency can be found on the net, factor that with the lowest signal level above the noise floor (and the closer the received signal is to the noise floor, the slower that data link is going to be).
Sorry to sound negative, but I think you will need line of sight at 2.4g, I am 99.9 % sure you will with laser, at this range.
Terry G7JFI
Merlin --- We're an autonomous collective... Help, Help, I'm being oppressed!!
It would be better to use satellite, cable or microwave to feed all but the smallest towns.
I wonder what the ultimate limit on baud rate would be, for example an analogue telephone line of 3KHz bandwidth manages about 52KB (NEVER 56!) on a V90 modem. That would equate to about 200MB on a 7MHz channel, enought to support a few users! It may be early days yet for this technology, with much better to come.
Nevertheless, an excellent achievement. It probably means that Mudamuckla and Bobo Creek will get broadband before I do. I live in an NTL cable area in London, have only waited for 3 years now, for them to upgrade the link from the street cabinet to wherever the internet connection is. Meanwhile, I am paying for digital TV that I don't really want, and a set-top box (actually a set-bottom box if we want to be pedantic, I have never yet seen one on top) with an RJ45 on the back which currently does nothing.