You can only go so far with simultaneous amplitude and phase modulation (FM is a form of PM). There are no other qualities to the signal. I doubt we can get many more symbols into a signal than the phone modem people have been able to cram in there.
Old AMPS, two different VHF allocations which I guess are going away.
800 MHz NAMPS.
GSM 900
GSM 1800
1900 MHz PCS, split into bands.
Satellite.
My understanding is that one of the GSM bands is allocated in the U.S., is that incorrect?
Regarding the cellular infrastructure, I am thinking of a market for "connections" that cell operators sell as a commodity to cell phone "networks", who would really be billing aggregators rather than networks. It really isn't necessary to build a whole network, you build a cell and make a wireline partnership, and you bill the aggregator per call.
Yes, you are right. But it has only been available for consumer use very recently. It was considered a military technology for a long time and even hams had to follow very strict regulations to use it, which only very recently have been relaxed.
Yes, the Europeans did a much better job on this. What I think is needed for this to happen in the U.S. is for business systems to be set up that allow the ownership of GSM infrastructure to be distributed among many vendors, with some central coordination to keep the vendors from interfering with each other. For example, I'd like to go into business to build a single cell, and get some portion of the connection fees from use of that cell. I'd also like to have some overlap so that cells can compete (and this does probably reduce frequency efficiency somewhat).
Some frequencies are efficient at heating body tissue. These include the frequencies used for cell phones. That is the major health problem. There is a good deal of regulation around it. Frequencies that are too low or too high generally stop at your skin. It's the ones that get through to your brain that are most dangerous, not because you'll pick them up in your thoughts (lots of us have tried but we can't make that work), but because heating that tissue might cause cancer.
The important thing to keep in mind is that propogation of energy follows an inverse square law. Every time you double your distance from the transmitter your exposure is 1/4 what it was before. Thus, a phone held up to your head is generally a much bigger risk of causing injurious heating than the ambient radio energy in your environment.
The question of how much bandwidth we can squeeze in per Hz is much thornier
To state it simplisticaly, you can get something less than 1/2 symbol per second per Hertz, and if you use phase for encoding, you can have more than two symbols, so this is more than 1/2 bit per second but in practice less than 15 bits per second.
The radio spectrum is a natural resource, nobody owns it.
Bands are a synthetic thing, what you actually want to know is how much bandwidth you can use. Essentially, we don't run out if we manage it well. The best way to manage it we know of so far is by using cellular techniques, which allow you to re-use the same spectrum every few miles, to connect wireless devices to the wired Internet. When spectrum gets tight, you build more cells, closer together, and reuse spectrum within smaller areas.
Where is the ceiling? Currently, it is defined by how high a frequency you can build an effective radio for. We can get into the milimeter waves, extremely high frequencies which theoreticaly contain much more bandwidth than we are using today. Current equipment for these frequencies is very primitive and tends to be wasteful of bandwidth, that will improve. Eventually we hit a ceiling defined by how well very-high-frequency radio propogates through objects - if it won't go through walls or windows, etc., its use may be limited to in-building use. There are also new technologies like spread-spectrum and ultrawideband that may allow us some additional frequency reuse.
The way the FCC is currently managing spectrum could be improved. They tried auctioning license rights off, and are still doing it, and this has resulted in 5 redundant bands for cellular phones, with about the same thing going on in each of those bands. If they'd worked out a way to better share the costs of the cellular infrastructure between vendors, we could have been doing the same thing in one band, building more cells as usage increased instead of adding more frequencies..
Again, not what I would write. Oracle, maybe, but didn't you notice that Interbase is going Open Source? In fact, they ran the license by me and it's really an Open Source license. Interbase is an enterprise-quality database and with the help of the community could become even more powerful. It would be nice to see participation like that from Oracle, but I won't hold my breath.
The problem is that the law currently doesn't recognize free software and places the same requirements on all software. IMO people who take money for software should be obligated to provide a money-back warranty. People who give software away should not be required to provide a warranty. The problem comes when somebody sells free software, as in a distribution. They can be made to pay customers back, and then can turn around and sue the developer to recover their damages. I want to protect that developer when it's free software.
I'm asking the attorney who brought this to my attention to write a paper. I'll attempt to get it published on Slashdot and elsewhere. That way, we can have an answer a bit more authorative than mine.
A possible scenario is that a distribution has to make refunds (or pay damages - there's a default warranty that may apply). Remember, the distribution is selling a copy of the software. The distribution can then sue the developer to recover its damages. Sure, Red Hat wouldn't do this but I would not put it past some of the others. We want that developer protected.
I'd be happy to exchange warranty protection for money. That's what most proprietary software publishers do as part of charging for their software. When I write free software, I don't provide a warranty, that is only fair. Anyone who wants a warranty can contract with LinuxCare or their local consultant to provide one, and pay for the privilege.
The problem is that UCITA requires warranties in some cases and prohibits contracts that waive the warranty. Then we have the question of whether or not our licenses are contracts, but we are on real thin ice there.
I am talking with an attorney who assures me there really is a problem. I can put you in touch with the attorney if you want some assurance that I know what I'm talking about.
This text from the GPL reads like a tear-open license, and thus a contract:
Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it.
It only made it to GCC after Stepstone made it clear that it wasn't interested in helping NeXT extend the language. So, NeXT did an end-run around Stepstone. That was much too late to help Objective-C succeed as a language.
Again, not what I would write. But you might get there someday. Please look up steganography and consider the implications for surreptitious exchange of data no matter what is used as a protocol. Perhaps you might want to lecture us on it a bit, too.
Want to bypass pattern-detection? Encrypt your connection. Blocking media by pattern detection can't possibly work over the long term. These folks have never heard of steganography, the science of encoding hidden data streams within other data. I can make any protocol look like HTTP, and in fact I can make any data look like valid HTML for purposes of exchange. Putting music in a GIF image is trivial, and you can even do it without changing the appearance of the picture.
But these manufacturers of bogus net filters will probably get some fools to invest in them. Maybe that's what the publicity is for.
If he were really Bruce Perens, he'd have a bit more understanding of transmission-line theory, RF, optoelectronics, what belongs in a point-of-presence and how the signal gets there, etc.
I'm sorry I couldn't stay longer. I had an 8PM flight, and a 2-month baby waiting at home.
I would be glad to come to Kansas again, and to other small, independent shows. All I ask the small guys for is travel and lodging expenses. If the show is well-financed and in an interesting place, I sometimes ask them to fly Val and the baby out to be with me. But I don't ask for any honoraria.
Slashdot Mirror
Thanks
Bruce
My understanding is that one of the GSM bands is allocated in the U.S., is that incorrect?
Regarding the cellular infrastructure, I am thinking of a market for "connections" that cell operators sell as a commodity to cell phone "networks", who would really be billing aggregators rather than networks. It really isn't necessary to build a whole network, you build a cell and make a wireline partnership, and you bill the aggregator per call.
Thanks
Bruce
Bruce
Thanks
Bruce
The important thing to keep in mind is that propogation of energy follows an inverse square law. Every time you double your distance from the transmitter your exposure is 1/4 what it was before. Thus, a phone held up to your head is generally a much bigger risk of causing injurious heating than the ambient radio energy in your environment.
Thanks
Bruce
To state it simplisticaly, you can get something less than 1/2 symbol per second per Hertz, and if you use phase for encoding, you can have more than two symbols, so this is more than 1/2 bit per second but in practice less than 15 bits per second.
The key is reuse, not carrying capacity.
Bruce
The radio spectrum is a natural resource, nobody owns it.
Bands are a synthetic thing, what you actually want to know is how much bandwidth you can use. Essentially, we don't run out if we manage it well. The best way to manage it we know of so far is by using cellular techniques, which allow you to re-use the same spectrum every few miles, to connect wireless devices to the wired Internet. When spectrum gets tight, you build more cells, closer together, and reuse spectrum within smaller areas.
Where is the ceiling? Currently, it is defined by how high a frequency you can build an effective radio for. We can get into the milimeter waves, extremely high frequencies which theoreticaly contain much more bandwidth than we are using today. Current equipment for these frequencies is very primitive and tends to be wasteful of bandwidth, that will improve. Eventually we hit a ceiling defined by how well very-high-frequency radio propogates through objects - if it won't go through walls or windows, etc., its use may be limited to in-building use. There are also new technologies like spread-spectrum and ultrawideband that may allow us some additional frequency reuse.
The way the FCC is currently managing spectrum could be improved. They tried auctioning license rights off, and are still doing it, and this has resulted in 5 redundant bands for cellular phones, with about the same thing going on in each of those bands. If they'd worked out a way to better share the costs of the cellular infrastructure between vendors, we could have been doing the same thing in one band, building more cells as usage increased instead of adding more frequencies. .
Thanks
Bruce (K6BP)
Bruce
Bruce
Again, not what I would write. Oracle, maybe, but didn't you notice that Interbase is going Open Source? In fact, they ran the license by me and it's really an Open Source license. Interbase is an enterprise-quality database and with the help of the community could become even more powerful. It would be nice to see participation like that from Oracle, but I won't hold my breath.
Bruce
Thanks
Bruce
P.S. I hope this posting comes out OK. My posting screen has "This box will be back once mu.current.nu is back to normal" all over it.
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
I am talking with an attorney who assures me there really is a problem. I can put you in touch with the attorney if you want some assurance that I know what I'm talking about.
Thanks
Bruce
There is legitimate reason for concern.
Thanks
Bruce
Bruce
Again, not what I would write. But you might get there someday. Please look up steganography and consider the implications for surreptitious exchange of data no matter what is used as a protocol. Perhaps you might want to lecture us on it a bit, too.
Thanks
Bruce
But these manufacturers of bogus net filters will probably get some fools to invest in them. Maybe that's what the publicity is for.
Bruce
Bruce
He's not me. Note the "." following the user name.
I would be glad to come to Kansas again, and to other small, independent shows. All I ask the small guys for is travel and lodging expenses. If the show is well-financed and in an interesting place, I sometimes ask them to fly Val and the baby out to be with me. But I don't ask for any honoraria.
Thanks
Bruce