Oof. Here I was just thinking about how cool it was to double-use existing infrastructure and to provide an alternative to the telco that didn't rely on renting bandwidth or wire from the telco you're competing with, and you come along to point out the stuff I should've thought of if I hadn't been distracted by the shiny side.
Hmm. Implementing structural changes to provide an electromagnetic solution won't be economically feasible because a big part of the reason for doing this in the first place is the ability to use copper that's already in place ... I wonder whether there's a mathematical solution to the interference problem. Obviously if you just ship the bits as unmodulated ripples added to the existing transmission line wave, you've got broadband emissions, and you can't do funky phase modulations to the base waveform because its frequency is (I assume) too low to be useful, but is there a useful encoding that would restrict the emissions to frequencies outside of radio communications use? Or is this one of those mathematical gotchas where we can prove that it's guaranteed to be a problem barring structural changes to the wiring (like adding shielding)?
As for eavesdropping ... ewww. Yeah, they're going to have to encrypt the packets, and that encryption is going to get thoroughly tested by black hats doing war-driving. Maybe some white-hats and grey-hats need to wander into the test area in MA and see whether they're currently at least attempting to encrypt.
Currently there are people who worry about eavesdropping using compromised routers (and other folks who put cash registers on WiFi without encryption <<shudder>>) When evesdropping on the connection becomes trivial, is it enough to personally run all your connections through SSH, or does the carrier also need to add encryption of the lowest protocols on top of that for you to be safe?
Oh man, I so don't want to wade through the math on a computer screen, but I know I'm not going to get ahold of it on paper any time soon.
Given my background, I wouldn't rely on SSH only. Unless I'm SSHing to a gateway somewhere, and making my other connections only via the proxy gateway. You see, I don't want to give any food even to the traffic analysis. I prefer to mask the IP addressess and port numbers as well.
I'm positive that a coding or modulation scheme can be concoted that leaves a small number of narrow windows in the RF spectrum less contaminated. A small number, narrow windows. As you no doubt are aware, the bandwidth occupied by a transmission is the symbol rate. So, if I have 2 Mbit/s line, it is going to occupy 2 MHz of bandwidth -- in segments, if necessary. Unless we get the center frequency up to 15..20 MHz, it is going to block everything. And the higher the frequency, the less optimal the transmission line, and the more it will radiate.
Power line frequency is not only far too low to provide usable bandwidth, but the various reactive and especially switching loads are continuously modulating the 50 Hz (or 60 Hz) carrier in an unpredictable and uncomfortable manner, in phase and amplitude alike.
The voluntary emergency communications infrastructure in the US as well as many other countries use uncoordinated frequencies: the parties are free to make their allocations within given bands.
What is going to happen is that the technology will be pushed through, despite the lone voices in the desert protesting against it. Only later when everyone and their cousin is relying on the technology, will the problems become evident, and a loud indignant noise will be heard saying "we didn't want this."
Speaking of lone voices in the desert, I have found it to require a bit of work to find objective expositions of any power-line communication technology. Either it is touted as the technology that will save us from any internet connectivity problems for the rest of time, or it is soundly denounced as the most devilish invention since [abomination of your choice].
Another problem with broadband on power lines is that it'll supply tons of *hash* directly to _anything_ that's plugged into a wall. That means you'll get noise on your stereo, DVD, TV, radio, and your _computer_. There's barely enough filtering in most consumer products to kill the sixty cycle hum - broadband will wreak everyone's home entertainment centers.
I'm not quite sure... Of course, there's an easy test. If the power cycling of the fridge, freezer, water heater, washing machine, and other such loads will cause an audible click or otherwise noticeable glitch in your equipment, then the data might be a problem, too. If, however, those high-intensity events are not witnessed on the equipment, I would feel the equipment would be immune to the data, as well.
My stereos do not pass this test. The computer does, as well as the FM monitor receiver. Admittedly, the last item is not strictly a consumer grade product.
But you are giving me hope. Perhaps, at a long last, we will see a benefit from the insufficient electromagnetic compatibility of consumer products.
(no subject)
Hmm. Implementing structural changes to provide an electromagnetic solution won't be economically feasible because a big part of the reason for doing this in the first place is the ability to use copper that's already in place ... I wonder whether there's a mathematical solution to the interference problem. Obviously if you just ship the bits as unmodulated ripples added to the existing transmission line wave, you've got broadband emissions, and you can't do funky phase modulations to the base waveform because its frequency is (I assume) too low to be useful, but is there a useful encoding that would restrict the emissions to frequencies outside of radio communications use? Or is this one of those mathematical gotchas where we can prove that it's guaranteed to be a problem barring structural changes to the wiring (like adding shielding)?
As for eavesdropping ... ewww. Yeah, they're going to have to encrypt the packets, and that encryption is going to get thoroughly tested by black hats doing war-driving. Maybe some white-hats and grey-hats need to wander into the test area in MA and see whether they're currently at least attempting to encrypt.
Currently there are people who worry about eavesdropping using compromised routers (and other folks who put cash registers on WiFi without encryption <<shudder>>) When evesdropping on the connection becomes trivial, is it enough to personally run all your connections through SSH, or does the carrier also need to add encryption of the lowest protocols on top of that for you to be safe?
Oh man, I so don't want to wade through the math on a computer screen, but I know I'm not going to get ahold of it on paper any time soon.
(no subject)
I'm positive that a coding or modulation scheme can be concoted that leaves a small number of narrow windows in the RF spectrum less contaminated. A small number, narrow windows. As you no doubt are aware, the bandwidth occupied by a transmission is the symbol rate. So, if I have 2 Mbit/s line, it is going to occupy 2 MHz of bandwidth -- in segments, if necessary. Unless we get the center frequency up to 15..20 MHz, it is going to block everything. And the higher the frequency, the less optimal the transmission line, and the more it will radiate.
Power line frequency is not only far too low to provide usable bandwidth, but the various reactive and especially switching loads are continuously modulating the 50 Hz (or 60 Hz) carrier in an unpredictable and uncomfortable manner, in phase and amplitude alike.
The voluntary emergency communications infrastructure in the US as well as many other countries use uncoordinated frequencies: the parties are free to make their allocations within given bands.
What is going to happen is that the technology will be pushed through, despite the lone voices in the desert protesting against it. Only later when everyone and their cousin is relying on the technology, will the problems become evident, and a loud indignant noise will be heard saying "we didn't want this."
Speaking of lone voices in the desert, I have found it to require a bit of work to find objective expositions of any power-line communication technology. Either it is touted as the technology that will save us from any internet connectivity problems for the rest of time, or it is soundly denounced as the most devilish invention since [abomination of your choice].
(no subject)
*That* should put the kybosh on it.
-m
(no subject)
My stereos do not pass this test. The computer does, as well as the FM monitor receiver. Admittedly, the last item is not strictly a consumer grade product.
But you are giving me hope. Perhaps, at a long last, we will see a benefit from the insufficient electromagnetic compatibility of consumer products.