backtop


Print 57 comment(s) - last by Neophyte1980.. on Nov 10 at 8:21 PM


Its not easy being green--Dr. Papandriopoulos has come up with technology which promises to revolutionize the internet.  (Source: The Sydney Morning Herald)
Australian researcher develops tech to raise internet speed x100, which should hit the market in 2 to 3 years.

It’s not easy being a young super-genius who just developed an algorithm to reduce interference in ADSL lines, potentially raising broadband connection speeds by a hundredfold. It’s not so easy, being internationally acclaimed and having fame and fortune knocking on your door.

Okay, so maybe life is going pretty good for 29-year-old University of Melbourne research fellow Dr. John Papandriopoulos -- but it’s not like he didn't earn it.

Dr. Papandriopoulos's PhD thesis is comprised with methods and techniques to use mathematic algorithms to reduce interference in DSL lines. This research is grabbing international attention due to the promise it holds; ADSL connections implementing the algorithm can go from approximately 1-10 Mbps to a blazing 100 Mbps or above.

Such an increase would revolutionize the internet industry worldwide. Dr. Papandriopoulos has two patents pending on his research, which will likely bring him some future wealth.

Further, he has been lured to move from his native Australia to Silicon Valley by Stanford University engineering professor John Cioffi, the "father of DSL" who helped develop the chips in the first DSL modems.

Professor Cioffi reviewed Dr Papandriopoulos's PhD thesis and was so impressed that he called up the young super-genius and offered him a job at his start-up company, ASSIA. Dr. Papandriopoulos accepted and is now on his way to fame and fortune abroad.

His Australian alma mater has not forgotten about Dr. Papandriopoulos either. University of Melbourne gave him the Chancellor's Prize for Excellence, a prestigious award. It’s no secret why the University of Melbourne is so happy; Dr. Papandriopoulos is assigning the University of Melbourne his patents, however; he will still likely make big royalties from the licensing agreement he reaches with the university. The university is eagerly looking into commercializing Dr. Papandriopoulos's technologies. They are in talks with multiple vendors of DSL equipment and modems.

Dr. Papandriopoulos's technology revolves around the interference and cross-talk that occurs between the cable wires of high-speed ADSL (Asymmetric DSL) wires. This significant cross talk and the resulting interference forces providers to settle for lower speeds.

"That is not an issue for voice calls these days but it becomes a problem when you're trying to wring more bandwidth out of these existing copper telephone wires [which power ADSL broadband connections],” explains Dr. Papandriopoulos. “This cross-talk in current day DSL networks effectively produces noise onto other lines, and this noise reduces the speed of your connection."

He says that his technology is better that competitive solutions as it is more practical and easier to implement. He sees it coming to the market with 2 to 3 years.

Meanwhile, Professor Cioffi and Silicon Valley can't wait to put the Aussie to work in only two weeks, developing more brilliant solutions. With a mind like his, perhaps the sad state of broadband in the U.S. and abroad can be overcome, and be replaced with ultra-fast and efficient high-speed connections.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

way to go
By Quiksel on 11/8/2007 9:04:47 AM , Rating: 2
WANT.

I do hope this actually comes to fruition. Around here, I can't recommend DSL to anyone, given the complexity of setup and poor speeds. Someone bringing speed enhancements of this magnitude would sure put the pressure back on cable companies to continue their development and increase speeds on their wire too.

This has to blow FiOS away, too, would it not?




RE: way to go
By therealnickdanger on 11/8/2007 9:31:19 AM , Rating: 5
I'm pretty sure that if this offers 100Mbps, it will crush FIOS. However, broadband companies will likely just withhold this technology to people that want to pay through the nose for it: upper crust, businesses, ubernerds.

BTW, the Enzo Ferarri is not supercharged, it's just an amazing NA V-12.


RE: way to go
By ted61 on 11/8/07, Rating: 0
RE: way to go
By LogicallyGenius on 11/10/2007 12:23:57 AM , Rating: 1
now the capitalist morons cant argue to charge more for research in internet speeds.


RE: way to go
By Polynikes on 11/8/2007 12:24:49 PM , Rating: 2
I got excited hearing about this, but you're right, this is capitalism; the telcos will have to charge more for faster speed. I'll pay for it if the price is right.

Sounds like this will make cable internet the new dial-up. Cheaper and a lot slower.


RE: way to go
By Oregonian2 on 11/8/2007 2:26:38 PM , Rating: 2
May be great stuff, but comments are a bit overdone.

1. There has been DSL technologies available that will do over 50-MBps for a long time. Just isn't implemented/used. Some only do it over shorter lines, but some are pretty fast over long lines as well.

2. Verizon is starting to install G-PON for FiOS. It's download speed is 2.4 Gbps and upload 1.2 Gbps and gets split up to 32 ways I think it was (it's split passively, the 'P' in PON so there's a limit on the number of splits). Remember that FiOS also is made to deliver a different HDTV signal channel to each house on the trunk, in addition to internet service and apparently other services they have in mind to roll out later. So they've some headroom -- but aren't going to cut their other revenue streams.

3. Advantage of copper pairs is that it's already there. Disadvantage is that often it's been there for a LONG time. The copper infrastructure apparently has relatively high maintenance compared to fiber. If it gets wet, it's bad -- fiber doesn't care. No crosstalk to even have to mathematically remove (and I suspect there are some assumptions in his algorithm that may not fully met in real systems).

4. All of these discussions are just the "last mile" link. From the central office (or not even there, perhaps in a neighborhood DSAM box) it's all shared resources (for internet service) and their OC-xx trunks probably won't be 100% dedicated to that one service (probably has the lowest QoS setting for that matter).


RE: way to go
By TwistyKat on 11/9/2007 10:04:04 AM , Rating: 2
For a second I thought you were referring to Enzo from Mainframe.


RE: way to go
By TomZ on 11/8/2007 9:39:49 AM , Rating: 3
quote:
This has to blow FiOS away, too, would it not?

The benefit of this invention is that it uses existing copper pairs that are already run to nearly every building in the country. FiOS, on the other hand, requires new, costly fiber runs. After all, those costs are the only thing that keeps us all from having fiber run to all of our homes and businesses today.

So yes, it could blow away FiOS, for that reason alone.


RE: way to go
By Blight AC on 11/8/2007 10:48:48 AM , Rating: 2
Well, it uses the existing copper... sure, but the network equipment connected to that copper will most likely need to be upgraded, and the current backbones will need to be upgraded as well to handle that kind of speed, so it's not a free speed boost for Telco's to give to their customers, but still cheaper then a Fiber Optic switch. Although Fiber Optic is the future anyhow, so Verizon is future proofing by going Fiber now. This new tech will help DSL providers in low profit areas (Rural and farm communities), where they can't afford to provide fiber to the door, to stay competitive.

And as far as blowing FiOS away.. I doubt it, once DSL providers get the infrastructure with the new technology to get to FiOS speeds, Verizon will just double FiOS speeds easily.


RE: way to go
By baseball43v3r on 11/8/2007 11:05:10 AM , Rating: 1
if i remember correctly werent the cable companies given like something along the lines (no pun intended) of 25 Billion dollars to run fiber throughout the company and then pass that savings on to the consumer? but like all things with the cable companies they pretend like that money never existed and that their service is just a courtesy and not a neccessity.


RE: way to go
By TomZ on 11/8/2007 11:25:20 AM , Rating: 2
Yes, of course the upgraded DSL would require cards in the central office (just like DSL does now), but those can be installed on a customer-by-customer basis (investment is immediately paid for), and it is relatively cheap to bring a high-speed link to the backbone into a central office.

The other problem with Fios is that it has relatively little coverage. Verizon projects they will have 7 million households covered (not actual customers) by 2010, out of a total of about 115 million households in the US. That means that just 6% of the US is covered by their service. And that is with a projected $18B in investment - that's "B" as in billion - which works out to about $2500/household - not cheap!

For the other 94%, i.e., the rest of us, a quick upgrade to next-gen DSL could be the ticket to Fios-type speeds.


RE: way to go
By Blight AC on 11/8/2007 11:48:23 AM , Rating: 2
Well right, for the majority of folks who don't have FiOS and only especially for those that only have DSL, this will be a boon.

I was mainly referring to where DSL is in competition with FiOS. While Verizon is paying through the nose to upgrade to FiOS now, it will payoff as the technology is pretty future proof.

Copper on the other hand is showing it's age, the new discovery will however, breathe new life into it. Course.. if they can get reliable 100 MB connections on Copper, perhaps it's not as obsolete as Verizon would like us to believe.. especially for the folks Verizon is providing FiOS too, as they are removing the old copper (to reuse or sell to reduce the overall cost of installing FiOS).


RE: way to go
By euclidean on 11/8/2007 12:05:18 PM , Rating: 2
quote:
The other problem with Fios is that it has relatively little coverage. Verizon projects they will have 7 million households covered (not actual customers) by 2010, out of a total of about 115 million households in the US. That means that just 6% of the US is covered by their service. And that is with a projected $18B in investment - that's "B" as in billion - which works out to about $2500/household - not cheap!


Ya, it's not Cheap, but they do still offer speeds for the same or cheaper price in all of the areas they're at. So it's very inticing for more people to jump on board, and with each new person, the total cost Verizon paid is going down bit by bit. I don't know about you but most people I know have been switching to FIOS and getting off of Comcast and Charter to be with a cheaper inet/tv solution. Also, FIOS literally is like a fire hydrant....depending on how much you want to let through you just have to open the valve more. They could easily uncap their speeds to 100mb/100mb and not even flinch. But if they did that now for the price they could get sued for trying to make a "Monopoly"...

But all in all, Verizon could you use this same technology on their existing Phone/DSL network and not even flinch. Hi capacity FIOS in town, Comparable speeds in rural/farm locations....I don't think this would hurt them much at all.


RE: way to go
By Screwballl on 11/8/2007 9:59:25 AM , Rating: 1
Remember, they are talking by 100x.. so in the case of DSL, this means increasing that 512k line to around 50Mb which puts it within competition level of cable for most areas. Cable has the capability of somewhere around 30-60Mb depending on location.


RE: way to go
By drebo on 11/8/2007 11:05:05 AM , Rating: 2
I'd like to know where the hell you live that your cable gets 30-60mbit.

For me, both cable and DSL cap out at 6mbit.


RE: way to go
By Korvon on 11/8/2007 11:29:43 AM , Rating: 3
The DOCSIS 3 spec (new cable modems) can reach speeds up to 160mbits. Now weather your cable company wants to give you that much or not is up to them. Current DOCSIS 2 modems max out at 30mbits. Shaw cable here in canada has a top tier service that is 25mbits. $99/month though.


RE: way to go
By drebo on 11/8/07, Rating: 0
RE: way to go
By wordsworm on 11/8/2007 12:01:42 PM , Rating: 1
quote:
This has to blow FiOS away, too, would it not?

I think it's on par with direct fiber. That's the speed they get in S. Korea anyways. DVD downloaded in 5 minutes anyone?

I'm certainly not on the same wavelength as this guy, but I can't help but wonder why the same theories wouldn't apply to FiOS as well.


RE: way to go
By ajfink on 11/8/2007 12:19:58 PM , Rating: 2
I'm pretty sure it has to deal with interference of the particular implementation of ADSL over copper. I don't think fiber has such problems. Either way, fiber is really only limited by the electronics on each side and what speeds they can deliver.

Makes me wonder about digital cable / cable network connections, as well.


RE: way to go
By Reflex on 11/8/2007 12:25:50 PM , Rating: 2
Because this is a method of reducing crosstalk, which is when frequencies in one copper line interfere with frequencies in another copper line in the same bundle. Fiber however does not use frequencies, it uses light pulses, and there is zero interference, so algorithms that reduce frequencies are not helpful in that scenerio since it solves a problem that already does not exist.


RE: way to go
By winterspan on 11/9/2007 3:02:52 AM , Rating: 2
I think we all know what you meant to say, but jesus.. you sound like a damn retard. I can't really comprehend how you ended up using the word "frequencies" for electrical current? Yes, that electrical current that is traveling over the copper pair is modulated at a certain frequency, but your use of the word is just plain odd. Alas, you are "correct", in that fiber optic cable does not suffer from the electrical interference /crosstalk/induction that copper lines do....


RE: way to go
By scrapsma54 on 11/8/2007 12:37:51 PM , Rating: 1
Airecomm is faster the FIOS. It's just expensive. It can go up to 60Mbps. Airecomm uses a satellite to transfer data. It has the same fidelity as XM and Sirus so it is certainly fast.


RE: way to go
By Etsp on 11/8/2007 1:12:51 PM , Rating: 2
If they use Geosynchronous satellites then the latencies are also at minimum 1/2 of a second, are probably dependent on weather, and possibly require costly fees to re-align in case of extreme weather conditions causing the antenna to move.


RE: way to go
By lumbergeek on 11/8/2007 1:26:14 PM , Rating: 2
Exactly. And it's not faster than FIOS. FIOS is deliberately limited by the ISP, it could easily do 10GB/s, it doesn't for commercial reasons not technical ones.


RE: way to go
By trisct on 11/8/2007 4:28:10 PM , Rating: 2
Not only that, but because of the way most TCP connections work the latency works against available bandwidth too. Unless the satellite company dedicates special hardware to correcting the situation for EACH customer, TCP-IP only allows 3 packets in-flight at the same time. With high latency that limits a socket connection over satellite you can only get about 200K per second, maximum, regardless of the channel size.


RE: way to go
By Etsp on 11/8/2007 10:28:56 PM , Rating: 2
Many satellite connections "cheat" by somehow pre-acknowledging packets. I'm not exactly sure how it works, it's just what I've been told. I work in a satellite internet company and I often see download speeds in excess of 1mb, so at least my particular company seems to have gotten around that limitation.


RE: way to go
By winterspan on 11/9/2007 3:09:09 AM , Rating: 2
Um... yeah he said 200K as in kilobytes = 1.6 Mbps. That's more than the highest speed I ever saw when I was using DirecPC/Hughes. And as he and others said, the latency is ABSOLUTELY MISERABLE. My average ping times were around 850ms-1000ms... totally rotten. It was unique back in the mid 90's when 400kbps down blew away a 56k modem. (the older systems were only 1-way.. had to use a dialup for the uplink)
When they went two way the uploads went to 150-300kbps but the latency was even worse. All in all, a pretty bad experience once cable and DSL service started rolling out in the late 90's and early 2000's. I wouldn't recommend it to ANYONE.


"Intel is investing heavily (think gazillions of dollars and bazillions of engineering man hours) in resources to create an Intel host controllers spec in order to speed time to market of the USB 3.0 technology." -- Intel blogger Nick Knupffer

Related Articles













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki