Slashdot Mirror
'Whispering' Wireless Internet
Zondar writes "MSNBC is reporting about a new radio filtering technology allows an ISP to use already-occupied frequencies to transmit and receive data. From the article: 'xMax, the latest innovation in broadband communications, is a very quiet radio system that uses radio channels already filled up with noisy pager or TV signals ...' and 'xMax is trespassing radio frequencies, although trespassing is not the right word, because we're allowed to transmit a signal if it doesn't interfere with other, stronger signals...' Too good to be true? Sounds like it would just raise the noise floor, to me."
This is extremely interesting, if not tried before. I wonder what FCC/ologopolies will have to say when someone else starts using their hard lobbied/bribed frequencies.
FP?
Live for the present, learn from the past, and dream of the future!
This is just some kind of spread-spectrum technology, nothing new... The signal consists of pseudo-noise. If the receiver knows the key to this pseudo-noise and can synchronize to it, he can decipher the message. This idea and this technology have been around for years.
from the article:
The first xMax network is currently being built in Miami and Fort Lauderdale where one base station can deliver broadband Internet over a 40 square mile area.
But with that much area, you need to start worrying about capacity. What good is it to cover 40 sqmi when you can't get a packet through:
The capacity of that wireless network is not bigger than any other wireless technology, which means that more base stations need to be added if a certain number of people are using the network -- typically several hundreds to a 1,000 users.
"Scientists don't change their minds, they just die." -- Max Planck
I didn't see any mention of the FCC in this article at all, something that may be indicative of a lack of approval from the relevant bodies. It's all very well the inventors/creators saying that this is technically ok, but when the people who are allocated the frequency range this technology operates in have problems with the raised noise or extra signals, or even just object to something else intruding on their licenced spectrum, I wonder what will happen.
Business Voyeur
I'm surprised that no-one's actually tried something like this before. What with the prevalence of radios that can adjust themselves to noise conditions, it seems that it would be fairly obvious to build one that could listen to the frequency (or frequencies) it wanted to transmit on and intelligently avoid stomping on other, old-fashioned signals in the vicinity. It's interesting, 'cause I just got done reading about something like this in this rather weird, but oddly compelling book.
Best quote from that article:
Instead, they quote the technology's inventor and the executive chairman of the company, while a man presented just as "an electrical engineering professor at Princeton University" actually sits on the company's board of advisors. None of these three, of course, have a vested interest in pumping up this new technology.
It's all the hassles of DSL but now with NO WIRES!
1000 transcievers over a 40 square mile area doesn't sound so bad to me. If the population is much denser than that, then wired net access is likely available.
An area of 40 square miles is a circle of radius 3.6 miles (5.8 km). Is that really more than a digital phone tower can manage, for example?
I’m old enough to remember 16K of memory being described as “whopping”
See this article for an explanation of some of the technical details of the system.
Mea navis aericumbens anguillis abundat
It isn't CDMA. It appears to be a combination of a narrowband pilot carrier and a wideband PPM signal that transports the data.
Mea navis aericumbens anguillis abundat
If everyone switches to wideband, low-power, densely-coded, mesh-network transmissions, then I suspect that the Earth will become virtually invisible to extraterrestrials who try to use SETI-style, pattern-in-RF methods. With nobody broadcasting at high power on a simple-coded narrow-band carrier, the RF emissions of the planet will become indistinguishable from noise.
I wonder if each civilization goes through a short RF-detectability phase before they so densely pack the spectrum with so many emitters that they become invisible, too.
Two wrongs don't make a right, but three lefts do.
The technical details are sparse but here are two links.
6 3700624 that has a little more information.
In the faq http://www.xgtechnology.com/faq.htm there is a brief description. Note that the spectrum plot shown is basically worthless because it does not show any signal details.
Here is a magazine article http://www.mwee.com/showArticle.jhtml?articleID=1
Note the following: In the first is the statement that Shannon's theorem is not violated but no justification is given. In the second it says that most power is put in the carrier. Both of these statements should raise red flags. Be skeptical!
If you know the characteristics of a signal exactly, you can recover it from below quite a bit of noise. One of the experiments I have my students do is to recover a signal 20 dB below the noise. It is trivially easy to do. The amount of data you can send, on the other hand, is approximately zilch.
Shannon's law describes the amount of data you can send as a function of the signal to noise ratio. As long as you are willing to put up with low bit rates it is no problem to use a signal below the noise floor.
Several of the posters have assumed that these guys have re-invented cdma. That's not necessarily the case (although it might be).
http://www.wirelessnetdesignline.com/howto/uwb/163 103775
An xMax-enabled system has several advantages of over a UWB network. Primarily, whereas UWB emissions require several gigahertz of spectrum, the "narrowband" version of xMax only requires sidebands on the order of several megahertz. The carrier synchronous nature of xMax also bests UWB, which uses thousands of pulses to represent one symbol.
Paradoxically, UWB is often designed as a PAN technology for use in the 3.1- to 10.6- GHz range and other limited uses in higher bands (24 GHz), leading to potentially high transmitter density. Given the amount of power emitted into adjacent bands, the cumulative likelihood of interference is high. In contrast, xMax is designed as a WAN technology, leading to a low transmitter density and lower interference potential. FCC rules also prohibit UWB applications from using spectrum below the 3.1-GHz band, whereas xMax is designed for sub-GHz use.
Lastly, xMax is a more efficient, agile system that requires as little as 6 MHz for broadband data transmission and can frequency-hop to vacant spectrum. As stated, the xMax signal is carrier-synchronous, making detection easier. UWB, on the other hand, doesn't use a carrier; timing must be embedded in the information, requiring large contiguous swaths of spectrum. Note that UWB requires higher signal power when measured using equivalent resolution bandwidth.
Here's an important clue, from their FAQ: "The narrowband channel allocation that xMax uses to coordinate reception of its wideband xG Flash Signal is not the system's information-bearing bandwidth."
So, it's a very narrowband pilot signal plus low level wideband signal with some new filtering/shaping tricks and maybe frequency agility on the wideband part.
The pilot is strong, easy to find, on a known frequency, shaped to occupy minimum bandwidth, and carries low-bitrate control info - like where and when to find the "flash" information-bearing carrier. It also may be a system clock reference (why not?). Being a clock reference would allow for more fancy demodulation techniques (yielding better BER performance) to be used on the other signal, because the lack of need to do clock recovery from the weak "flash" carrier.
The point is lower power. Since signal decreases as a square of distance, even small reductions in transmit power will have a dramatic difference in the noise signature of the Earth at multi-light-year distances. Ultra wideband allows lower power.
As an aside, the transition to heavily encoded packet RF also reduces our signature to ET. Anyone with a long enough wire and a speaker can pick up analog TV or radio and recognize it as synthetic. Can the same be said for highly dense encrypted digital traffic? Even my 56k modem sounds like white noise to me.
Build a man a fire, he's warm for one night. Set him on fire, and he's warm for the rest of his life.
although trespassing is not the right word, because we're allowed to transmit a signal if it doesn't interfere with other, stronger signals...
Damn right it's not the right word, and it wouldn't be even if it weren't legal to transmit on those same freqs. You can't trespass on frequencies because frequencies are not anyone's property. We gotta shake off this relentless trend of treating rights and licenses as property. To use a more familiar example, nobody "owns" music, not even the composer. Rights holders don't own anything at all, they merely control the rights to do specific things for a limited time.
The distinction isn't semantic nitpicking, it's very important because treating rights as property gives the copyright control industry an unfair advantage in any public discussions about rights issues. They like to play the part of the plucky little old lady chasing down a purse snatcher, or the outraged homeowner defending his castle against burglars and government goons. They get away with it because the public has been taught to overlay the simple and familiar concept of property on much more complicated issues. Treat rights as what they are -- temporary conditions set by the government -- and various rights and DRM issues suddenly require a lot more thought, which they should.