Slashdot Mirror
Aussie Claims Copper Broadband now 200x Faster
SkiifGeek writes "Winner of Melbourne University's Chancellor's Prize for Excellence, Dr John Papandriopoulos could soon find himself the focus of a number of networking companies and government agencies interested in wringing more performance from existing network infrastructure. Dr John developed a set of algorithms (US and Aussie patents pending) that reduce the impact of cross talk on data streams sharing the same physical copper line, taking less than a year to achieve the breakthrough. It is claimed that the algorithms can produce up to 200x improvement over existing copper broadband performance (quoted as being between one and 25 mbit/sec), with up to 200 mbit/sec apparently being deliverable. If the mathematical theories are within even an order of magnitude of the actual gains achieved, Dr John's work is likely to have widespread implications for future bandwidth availability across the globe."
(Up to 200 mbit/sec) / (Up to 25 mbit/sec) = 8x improvement...
Search RapidShare and MegaUpload!
Exactly. Without intending offense to Dr. Papandriopoulos, this is really not news, nor does it have widespread implications for future bandwidth availability across the globe. Global bandwidth is more about high speed backbones, which this technology does not even begin to approach. It is only useful in solving the last mile problem of getting things off the backbone to a terminal. And by the time this gets commercialized, I think we can count on at least three other technologies being faster still, with cellular style broadband probably at the top of the list.
Had a few beers with him. Here is his homepage.
Nope.
I can bet that it is a reuse of the 3G MAC ideas. 3G uses multipath to improve the signal to noise ratio by filtering the signal versus delayed samples.
Similar thing is possible with crosstalk as long as you handle all wires from the same duct in the same ASIC this usually is not the case. It will simply not work in countries where access to the copper is unbundled. In other places it will require major rewiring in the exchange.
I would hate to extinguish the hopes of all hopefuls which think that the holy grail has arrived. This type of algorithms provide O(LOG N) improvement and there is major improvement only for the first couple of filter buckets. Once you are past that each bucket adds less and less.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
I know this guy though having attended conferences with him. I know he is not a scam artist. I also think he is brainy enough to do this. He is not a fly by nighter but a serious communications theory researcher with a track record. As I've just emailed to my supervisor, "It's not every day a communications theorist makes the mainstream media". John Papandriopoulos is easy to find on google.
The slashdot summary and linked articles are rather short on details. A little googling located some details:
NOTE: I did a quick skim of it and had not seen any empirical evidence of the advance; seems to be entirely theoretical. I don't mean to lessen his accomplishments, but my experience is that reality usually has unforeseen factors. I certainly hope he IS on to something here!!
(*) I didn't know anyone used the <blink> tag any more. :/
I think the premise that this tech is based on 3G multicast is wrong too.
Dr Papandriopoulos paper suggests the algorithm works by iteratively lowering power, and therefore reducing crosstalk. The reduced crosstalk allows faster protocols like VDSL to be used on the copper that was previously only capable of ADSL2.
"I've got more toys than Teruhisa Kitahara."
You underestimate the cost of replacing the last mile technology... there are millions of miles of copper out there and it isn't going anywhere soon. BT's 21cn replacment for example is going to take until 2011 to update their network (if on schedule, and AFAIK it's behind already), cost many hundreds of millions and *still* relies on copper for the last mile (it merely makes ADSL2 deployment easier). And most countries' networks aren't even coming close to that level of investment.
If this means they'll be able to go to ADSL3 at 200Mb/s then I'm all for it.
A short one: Yes and No. It still stands, the numbers are still correct. That's the theoretical limit if you use the normal phone exchange(s), and the existing, limited, phone bandwidth (300-3400 Hz)
ADSL, though, uses the spectrum above, and needs extra ports on the last phone exchange to your house, since - contrary to standard modem - these signals don't pass through the plain old telephone system. They are kind of injected at the very end.
Comment removed based on user account deletion
Adams platform.
Also Australian. Who would have guessed it's an island full of criminals?
Umm... a wire wirelessly push a signal? Lets use the unusual and very exotic term, induction. That means wires wirelessly push signals onto other wires. We could use back emf, too, though that is a little better known. This is a tech site after all.
Semi-automatic amateur armchair Australian philosopher; conjecture ready at any moment...
It would travel to the break in the line, hit the end, and travel back towards the source
all discontinuities cause some degree of reflection and it can be a big issue as frequencies get higher. Telco wiring is likely to be full of discontinuities (cross connect panels, different cable types etc).
destroying everything in it's path.
Luckilly it doesn't destroy everything in it's path. It destroys some frequencies attenuates others and boosts others. Oh and it causes some nasty phase effects too. It is a very similar effect to that of multipath distortion in radio systems. The fact that primitive systems like thinnet couldn't cope with this doesn't mean it is impossible to do so.
What has really changed (and continues to change) is the systems we can put on the end of a line. DSP chips get ever more powerfull and with them ever more complex encoding schemes become availible. Systems can split the availible bandwidth into narrow bands an then tailor the encoding perameters to match what is going on in each band.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register