Slashdot Mirror
Massive Bandwidth over Powergrids?
The LoneSmurf wrote in
to send us a news.com story that talks about soem company's
new technology that they claim will allow gigabits of bandwidth
to any outlet in your home.
The article talks about the skepticism, but there really isn't
much technical stuff in there for us to consider. It sure
would be great, although I'm not holding my breath.
I don't know much about this, but I think they're
only planning to cover the distance between the local powerstation and your house. The powerstation would then be connected to the Internet through normal means, making it kind of
your ISP.
Connecting a powerstation to the internet through a special cable is quite cheap, connecting hundreds of appartments using the same technique is not, so there are some great savings to be made.
/Tord
Here's more broad-band hype. The only people who get the benefits of broadband are the first adopters. As the use of broadband increases, the Fourier transform shows us that the noise floor increases.
Bandwidth in ANY medium is intrinsically limited. TANSTAAFL.
Those amateurish diagrams on MediaFusion's web page; the utter drivel standing in for non-existent technical information; natter about "inscribing" the data onto the low-frequency AC; the complete doubletalk crap about "modulating [not their word] the magnetic field" as if it were possible to manipulate that independently of the megawatts of 60 cycle; the bibliography clearly stuffed with impressive sounding titles culled from random sources; and that's only a few things that come to mind.
... where was I? Oh, yes: if this isn't a hoax, where "hoax" includes falling short of the fantastic bandwidth claims, then it would be pretty exciting - nearly as much so as The Phantom Money-maker has been. But I have the strongest suspicion that, like TPM, it will turn out to be so much simulation and special effects. If there were a real revolution here they wouldn't have a picture of their network that look so very much like a bad imitation of "how X10 can automate your home appliances."
If this isn't a hoax - and hyping a technology similar to one that has already been developed and set aside because it falls drastically short of the bandwidth needed to be useful for a large number of connections (MF shows their "head-end" interface connected to a substation - oh, and notice that they don't show their fabulous technology being used for the uplink from there - that's exciting new microwave to the local telco CO)
Really, it's hard to choose what to criticize here: it's all so utterly bilge from end to end.
I give it a 0 on the plausibility scale, and then take some points away for being two months behind its rightful time.
I'm posting AC b/c I'm not sure if this'll
get me in trouble or not. I think it is common
knowledge but I could be mistaken.
I work for NT and one problem they ran into in the
UK tests of this technology is the fact that any
large metal objects that the power lines were
connected to (i.e. street lights, etc) started
acting like huge radio transmitters, broadcasting
the data everywhere.
You miss the obvious -- that all this excess radiation is going to cause interference all over the place. It's bad enough that a 50/60Hz hum and associated harmonics are always showing up in wires everywhere due to the lack of shielding in powerlines and house wiring, but now you want to spew even more RMI throughout the spectrum with your networking-over-powerlines hack?
Powerlines were never designed to carry bandwidths required for transmitting data signals. As such, they lack fundamental safeguards against RMI such as shielding and are thus unsuitable for data networking.
NetBSD: the cathedral vs the bizzare.
Not only would shielding be a problem, but what about the sharp spikes created by light dimmers? Those SCR's turn full on anywhere in the 60 Hz cycle (they are great noise machines for AM radios.)
At work, we have hundreds of motor controllers up to 500,000 watts that make an oscilloscope on the mains light up due to noise (and that's after the snubbers AND isolation transformers.) So, there might be a problem with communications in many places...
You might not be able to switch at that speed, but you can modulate on multiple lower frequencies.
This makes me wonder why consumers just can't use microwaves for high bandwidth transmission. There is a lot of bandwidth and an amazing amount of channels at 2GHz and above.
To minimize cross traffic in high density areas, they are line of sight. Depending on the antenna, they can be sharply focused at a target or in a broad beam.
I have seen X-band door openers for low as $20. Surely, these can cheap enough for consumer use.
For people afraid of microwaves turning people green with colorful cancerous bumps out of your head, just think about the energy level of a 1 watt microwave transmitter that can be used for a computer versus a 100,000 watt radio station transmiter. The telephone company uses microwaves for relay stations, but I doubt they want this to catch on...
In the US, most places I have seen have a transformer on every street corner. Its because great savings on wire and current losses can be realized with long skinny wires to a 13,500 volt transformer. A skinny wire to your block and a big, fat cable to your house is cost effective. Transformers have other benefits as well, like absorbing much of direct lightning hits.
Promise the world and watch that stock price jump. +4.44 points? Look at some the claims:
... the company has said individual consumers could get network connections of 2.5 gigabits per second--an estimate the company calls highly conservative ... But even at that speed, one could download the entire contents of an average computer's hard drive in a second
2.5 Gigabit is a long ways away from a 6.5 gigabyte hard drive.
Anyhow, I suppose you could have a high bandwidth transmission over the power lines, even without the high frequencies. Just imagine the transmission as a current loop with the current modulated at extreme levels. Unfortunately, the cost considerations to do this seem very interesting.
Hydro-Quebec tested this idea on their power lines. The original idea was to use their lines as a connection between their different (and usually far-apart) plants.
After a successfull technological trial, they thought they could actually serve the entire province and split off an ISP branch.
However, plans did not go through for a couple of reasons, at that time (this was 5-7 years ago):
a) the transiant background noise would make it impractical for most purposes
b) the CRTC (broadcast gestapo of Canada) did not grant Hydro-Quebec with a permit to broadcast signal, citing unfair advantage.
The second reason would probably be moot today, as their tigh... regulatory board, loosened up a bit in the last couple of years.
The main problem with this is that power lines are not nearly as well-balanced as the twisted-pair that feeds your phone. The result is that any signal on them tends to be radiated, and the FCC takes a poor view of this as would any radio users.
Even DSL is a problem in this way - the phone lines were not intended to be used at DSL frequencies and the result is often broadband radio interference where DSL is used. The problem is much worse for power lines.
There's also the problem that the transformers probably won't pass high frequencies too well, and someone will have to put bypass capacitors on every one of them to get the signal around the transformer.
By the way, here in North America we use 60 Hz, while Europeans use 400 Hz. Another poster had that backwards. 400 Hz transformers can be lighter and smaller than 60 Hz ones, but you probably still have a lot of them (even if you don't notice them).
I want switched fiber to the home. Maybe in a decade.
Thanks Bruce Perens
Bruce Perens.
I wont comment on all of your post, but living 1mile from the norweb trials of net over powerlines test, i can safely say for the 1st 2 weeks of the test, the street lights most certainly did ficker, though it was later rectified.
Um, how much electricity do _you_ use?
1 kilowatt is a lot of juice for one square meter... I could get over a hundred of those
on my roof... I could probably power the whole
neighborhood with that kind of power. The max
you could run into a home is 44 kW, given 220 volts with standard 200 amp cable coming in.
And nobody's going to be using 44 kW unless they're growing marijuana in the attic or something.
Of course, 100% efficiency isn't going to happen.
And the ugliest solar panel array I've seen is
still better looking than the powerlines strung
all over the place... maybe it's different in
your town. But newer designs are incorporating
the solar cells into roofing shingles, so they
aren't that bad looking.
And someone mentioned batteries being a disposal
problem- right! Flywheels are the answer, as soon
as they can develop the right composites that
don't explode when they break at 10000 rpms.
There's been alot of comments on bypassing the transformers. We already know what the frequency is of what we want to block - 60hz. Why not simply use a low-pass or bandpass filter and then throw a wire across the transformer. That would eliminate the interference caused by the transformer.
Now you need to figure out the attenuation problems (high frequency = high attenuation).
--
This is available *now* in the UK and as far as I know the rest of Europe. It's really not new technology at all, though I think it was easier for them to implement overseas because they use a lower frequency current. I don't recall who offers it but I'll find it and repost...
Problem is that while it's possible to provide high-bandswidth for home countries, how would it help with misery trans-atlantic link? What if your super-duper VideoPhone works for US/Canada and you can't call your friend in the UK?
However, we need cheap high speeds anyway and this is a viable option. Problem is that telco's don't really want this happen as it would ruin their business where they charge people for (previously) limited resource (bandswidth). They should face that in 21st century (which is right behind the corner) bandswidth will be right and not priviledge. The only major difference between humans and other's repsesentative of the Mother Nature is ability to COMMUNICATE EFFECTIVELY.
AtW,
http://www.investigatio.com
alexc
Join Majestic-12 Distributed Search Engine
Nortel has done testing in the UK for these type of communicatons.
As I recall they took advantage of the fact that most powercables there run underground, so they are pretty well shielded from the start.
As someone else pointed out, you can't get the signal past the nearest transformer. The point is that most newer transformerstations already have optic fibre installed. Most underground cableinstallations today lay down optic fibre together with the power cables out of foresight.
/olle
A lot of posts on this board seem to think that sending network signals over powerline is a brand new idea, or that it is impossible or requires special insulation on the wires. The company I work for (Intelogis) has had a product out for nearly a year now that is a cheap, albiet slow, powerline network.
There are a few problems with powerline networking that people have brought up. I'm just a software guy at this company, but I'll try and address some issues that have been brought up here.
1) It can't go over transformers
Well, no. A transformer is the physical limit of any kind of powerline network, since it gibbers up the signal so much. Powerline can be used to distribute broadband once it reaches the home, but it can't carry the signal TO the home.
2) Nearby street lights broadcast the signal
Um... this is just plain silly. But the signal can be snooped by your neighbors who share your transformer. So we encrypt the data. Problem solved.
3) It's too slow
Yes. Our current product only runs at 350K, making it a bad solution for technology shops. It's primarily aimed at the home office or the small office. We don't use it ourselves here at work since we have bigger bandwidth needs than that. But we are going to be releasing a 2 megabit product later this year, and hopefully a 10 megabit product soon. That's nowhere near the gigabit range this company in the article is claiming. Personally, I think it's all hype. They'll milk their shareholders for a year or two, and then call it quits. (Our company, on the other hand, has had a working example of a powerline network on store shelves for a year.)
4) Light dimmers will spike the signal.
A lot of things will cause the network to drop packets. Our current product will detect that and just re-send whatever packets were dropped, same as any other network protocol. Our next product has a lot more redundancy built into it; we send multiple signals at different frequencies, and they don't all get disrupted at once.
5) There is a lot of bandwidth in the power grid that we could be using
Yes. But to send a signal over powerline, you have to send it at such a high frequency to avoid interference over the wire that the signal tends to bleed off. (I don't understand all the physics behind it myself...) Power lines are noisy, but they can carry some signal. They are not practical for connecting a whole city block to the Internet. Powerlines really are not a good solution for getting broadband to your home. But power companies could use other ways to get a broadband signal at least to the transformer and from there pump it into the home with powerline.
6) Star Wars Episode I isn't making any money.
Ha! Star Wars is cleaning up at the box office right now. Where did you read this? CNN or some other similar trash network?
If anyone's interested in checking out our product, we've open sourced our drivers. You can get them at:
http://www.intelogis.com/opensource/
(Render sits back and waits for the flames...)
Fiber has the wonderful advantage of being extremely easy to shield. It's a lot harder to do that to electrical cables. Also, the ultimate bandwidth available on fiber and other optical carriers is huge. Electrical systems are much more limited (shielding is difficult, parasitic capacitance and inductance are big problems, and the electrical characteristics of the wire itself get ugly at very high frequencies).
And this isn't about transmitting over something like cable, which is shielded and optimized for data transfer - it's about transmitting over household wiring. Household wiring is unshielded and has very messy geometry. It also has a huge number of devices with wildly varying electrical characteristics hanging off of it. Unless you're working under ideal conditions, trying doing serious data transfer over it would produce a garbled mess.
Phone lines within a house are a bit better, as there's less crud hanging off of them. Shielding still isn't great. 10 megabit has been achieved in commercial products, here.
As far as transmitting high frequency data over a city's power grid, good luck. Reflections, interference, and cross-talk will be nightmares.
We'll need a fiber infrastructure at some point anyways. I say just push for it instead of trying to use existing copper networks.
Well, I think this is not only a good idea, but one which is easily possible. If you have ever scoped AC signals before, you'll know that there is a lot of noise at a lot of high frequencies.
It should be trivial to harness some of the higher bands for transmitting data. The only problem is that power cables used to transmit the data would probably have to be extensively shielded, as they would become transmitters and receivers, either losing or broadcasting the data signals at points. Normally, as the AC single is just a 60Hz signal, power cables are often just shielded to protect against the elements, not against transmission of signals.
æeee!
While I certainly hope this company's claims are true--I wouldn't mind a gigabit pipe into my home--those promising `virtually unlimited' bandwidth should be viewed with the same skepticism those who promised virtually unlimited energy from a cold cup of water were in the last decade. Let them prove their work--perhaps write an article for a scientific journal detailing their work (they may patent their work if they need to keep it a secret)[1].
Cheers,
Joshua.
[1] This is not intended to be an endorsement of the current patent situation.
--jon. Postel is dead. May we all mourn his, and our, loss.
Yes you are certifiably paranoid.
-----BEGIN OFFTOPIC-----
That said, any "Big Brother" organization that wanted to tap into many homes and businesses would logically turn to power lines as a data transmission medium. The trick is to have consumer devices that transmit data without the designers knowing about it (forgetting about the problem with Transformers for the moment). I can think of very few ways that "Big Brother" could surveil so many people in their homes. Other possibilities inlcude the phone network and of course Echelon.
Each of these would require the complicity of many organizations, hardware and software techs/vendors, repair techs, and others. While an automated system, that is not actually connected to the rest of the world, such as Echelon can hear much--complete and total omniscientness(sp? I made this up) is still too manpower intensive and requires too many people who know about it. Remember: The ability to keep a secret is inversly proportional to the number of people who know about it.
The intelligence industry is composed of individuals, just like you and me, they are not perfect and not all-knowing. Infact many are incompetant, good, moral, evil, power-hungry, insecure, long-hair, crew-cut, and have wife, family, sister, father, new Camaro, and are just like you.
-----END OFFTOPIC-----
-- Remember: Wherever you go, there you are!
I attended some meetings with german companies and cooperations with foreign firm (uk, us, canada), who are testing digital powerline in germany.
At the moment they are testing it under real life conditions and get speed at about 1 Mbit/s at the max. Under controlled enviroment they achieve about 2 MBit/s and they're talking about 10 Mbit/s. I don't think they'll get 2 Mbit/s or even 1 Mbit/s for the average user (here in germany about 200 househoulds share a connection to an electric transformator and so share the maximum connection speed. Perhaps you'll get about 256kbit/s at the moment (because at the beginning not so many people use this technology)).
I don't think that at the moment gigabit-speed is realistic because you don't have a controlled enviroment. Every household has extended it's powerline (just thinking about the wires in my student flat....) .
I like to have a high speed internet access in my room (with the digital powerline, adsl...who cares), but while listening to the marketing people at some firms (for example the Bewag, the Berlin Electric Company who is testing the digital powerline) I get a strange feeling. They aren't talking about high speed internet access or an alternativ to access the plain old telephone network. They're talking about new services like securing your home or your car (they want to use devices in the street lights to stay in contact with your car and detect unauthorized movement of it....imagine: someone call's you: your car is stollen, but we already got the thief.....). It sounds a little bit like big brother is watching you to me...
Bye,
Keef (enjoying his 10 Mbit/s Connection over a Laser Link)
but my Step-mother's cousin is developing at his company that sounds similar to this story.
Apparently there is a whole mess of EM "space" in electrical power lines that are not used for anything... only a small portion is actually used to carry electricity. So, my cousins' company is developing a way so that Power companies can use that "space" to transmit data about the usage/time etc. Obviously, this is really good for places with very remote communities that you can't get a meter reader to.
They had a two or three provinces in China very interested in the technology...
Anyway, I seem to remember him saying something along the lines that there is a whole lot of bandwidth in the Power Grid that we could be using... it'd be cheap too.;)
This reminds me of the pulse radio technology.
Measured before on slashdot.org and FreakTech
Maybe they have combined the pulse radio technology in a power grid modem.
The companies claim of gigabit/s is far faster that the europaen experiments, whish have transferee speed around megabit/s.
Knud
O.K. so it may be a good idea, but isn't anyone a little scared of plugging 110V AC into anything other than your power-supply.
if your AC-NIC fails, and sends that current into your system you're gonna be replacing alot of pieces.
-or worse if it send it out over your LAN.
"Nyquil - The stuffy, sneezy, why-the-hell-is-the-room-spinning medicine."
With all the hullabaloo surrounding Echelon recently, how do we know that such a system is not already in place? Think about it:
That new toaster you bought with the cool digital lightness/darkness controller in it... Is it tapped? The element inside could be adapted into a crude microphone, picking up conversations and feeding them to voice recognition Automata in the controller. The controller could screen for potentially illicit keywords, like "Bomb" or "Tinky-Winky". The conversation could then be broadcast to an Echelon archiving site through the local power grid.
And why not a toaster? In every civilized nation on the planet, the truly important conversations happen in the kitchen. The reason is simple, really - that's where the food is.
Unplug your toaster NOW!
=)
If you think this (my sense of humor) is disturbing, think about how easy this would be to implement.
http://news.bbc.co.uk/hi/english/sci/tech/newsid_3 54000/354646.stm
A demo house in the UK (not known for its sunny days) which has produced 50% more electricity than it has used.
http://www.pv.unsw.edu.au/
World record holders for PV cell efficiency - 25%
Generally the idea is that you remain connected to the grid and any excess energy produced (like during the day when you are at work and dont use lights or the TV etc) gets fed into the grid for 'storage' (actually used at places that need it) and then you suck it back out at night when you are at home and you need it. This means that power stations don't need to be as big and are mainly there to handle the peaks and the off hours.
Of course, having an energy efficient building and appliances goes a long way to help.
And PV cells just need light to operate - so even when it is cloudy, they are still generating, just not as much as with direct sunlight.
.
MrCreosote Meow!Thump!Meow!Thump!Meow!Thump! "You're right! There isn't enough room to swing a cat in here!"
It can be implemented easier in Europe because they commonly have more houses served from each power transformer. In North America (Canada, at least) there are often only four or five households served by each step-down transformer on the pole. The signal would have to be injected at the secondary side of the transformer since I can't imagine high speed data passing through a big ol' iron-core transformer optimized for 60 Hz.
Go read some past issues of NTK (www.ntk.net), they point this out as a media hoax.
But the hoax was based on some legitimate concerns about re-radiating your private information along all the power lines to the sub-station, which NorTel couldn't answer.
I've got to go post a top-level comment on this...
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
In parts of the US and Canada, and most places in Europe, electric meters are being replaced with newer models that can be read on a regular basis from collection centers.
The places where I am familiar with the technology the electric companies have had to place small bypass filters at each transformer and switching station, but the costs are small compared to the cost of sending a live person around every few months.
I've seen the signals on 8.something MHz, as well as 1.7 MHz. Every meter has a unique identifier similar to a MAC address on an ethernet card, and gets polled on a regular basis. This also allows the company to detect fraud and out-of-bounds usage on a near real-time basis.
The machines which collect the data are typically located in sub-stations which supply 10,000 to 35,000 customers. They can poll each customer in a sub-zone about once every 2 days, and its a continuous process. They are small unix minis, some new ones use NT, and they feed the data directly to the billing center.
If you poke around on the web, you will find that some German hacking groups have documented the protocol used, its pretty simple. Haven't heard of anyone inserting their own information back onto the line or causing DoS attacks. I would expect the companies that make these products also have a web presence.
the AntiCypher
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
There is gigabit ethernet going here, both over fiber and copper. Requires 4 twisted pair wires to get the signal to 100 Meters on Cat 5 cable. Terabit/sec experiments are underway for products to hit the shelves in 4-6 years. Exabits? I smell a media hack!
The biggest problem with the NorTel experiment in England was that power lines are almost never twisted, just straight parallel wires. Even though the transmission used a highly redundant coding scheme of something like 17 of 5 bits (85% of the information is redundancy, 15% is the actual transmission rate), there was problems of noise and crosstalk. Then there was still problems with putting enough power into the return signal to get it back to a sub-station intact. When they did that, the signal was detectable at every other power outlet on the sub-grid. There was a good media hack done in England which put egg on the faces of the NorTel team, since they couldn't clearly deny the charges of radiating peoples private data all over a region.
So where do you put your firewall when every one of your outlets is an internet connection? What's to stop a blackhat from plugging into an outdoor light socket and cracking every house on the block? Sounds like a whole new field of hacking/cracking just waiting to be exploited.
But then, maybe everyone should just be using fully encrypted and authenticated IPv6 between every device in their house.
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
This kind of stuff already exists in the form of X-10 computer interfaces. The solution is to use optoisloators. These are basically IR LED/detector combos facing each other in a sealed plastic shell. You mount one of these on a board, and if the high voltage portion of the circut fries, the IR LED burns out, but no high voltage signal leaks to the other side. You can use things like fuses too, but those aren't 100% reliable especally for quick spikes.
C'mon guys. Don't you recognize a scam when you see it. The power companies have been trying for DECADES to find a way to get information to flow over the power grids(I know because I was working on this problem for a power company in 1979). They could save 50% of their operating budget if they didn't have to visit your house to read your meter. OF COURSE you can send a signal over a copper wire, DUHHH!!! This is NEWS? But high frequency (multiple 100's of Mhz and up) signals won't stay on an unsheilded wire, it just acts like an antenna. And even if the wires themselves weren't unsheilded noise-ridden super-antennas, there's always the problem of the TRANSFORMERS. Huge blocks of metal with big coils around them. They are an incredibly efficient high-frequency blocking filter. The high-frequency information bandwidth through a transformer is precisely ZERO point ZERO bits per second. Another tip-off that the article is either a JOKE or a SCAM is the data rate that is mentioned. Exo-bits? Ten to the eighteenth power bits/sec??? That's a frequency somewhere between ultra-violet light and X-rays. You can't send that kind of signal over a copper wire, it would boil the electrons off the atoms. Don't they teach even rudimentary Physics in our schools anymore?