Patents get issued all the time for nonsense, and things that do not work. A patent that I got, we proved (after filing) that it did not funtion as described. Two years later the patent still issued. Go figure...
If this gets challenged in court it will fall apart. Too much prior art. I would start with Morse Code...
After reading the information here (and the limited FCC stuff...) I believe that the term "WiFi" is being used incorrectly here.
Allocation of spectrum will lead to use, but the 802.11 protocalls are not optimized to this type of distance broadcasting on allocated spectrum.
My guess would be that 802.16a (a.k.a. "WiMax") might get used here, although the original frequency definitions for this were from 2-11GHz.
With the allocation of spectrum down under 700MHz, the RF front end definition of 802.16a, could be modified to go down to these bands without a lot of pain.
WiMax makes more sense here. Orderly calling between subscriber station and server tower, rather than the "bumper cars" mentality of WiFi. It is optimized to a multi-user environment.
WiFi was never designed for long distance communication or heavy multi-user environments. It amazes me that people attempt to use it for such. But then half the fun is seeing what you can do with it, rather than what it was actually designed for.
News for nerds..... Obviously not communications nerds....
A MAC layer in communications terminology is the digital protocalls involved with all the control functions, ergo, priority, identity, and similar.
A PHY layer is the "physical layer" that is the RF frequencies involved, the antenna, the RF bandwidth of the channel and similar.
802.15.4 is designed for really low duty cycle use. Turns on once in a while, does its thing briefly, and then turns off. Think thermostats without a wire connection, or a set of burgalar alarm sensors without wires back to the alarm system.
Because of the low duty cycle it lends itself to battery power pretty well.
The Zigbee alliance is here: http://www.zigbee.org
And the IEEE standards committee behind this effort is here: http://www.ieee802.org/15/pub/TG4.html
Actually getting the MAC and PHY layers (now that you know what they are...) onto a single chip is pretty good thing. Most groups split it into a big digital ASIC (the MAC layer) and then an RF front end chip (the PHY layer)
Suggested option, put the metal that the magnets are glued to in a vise, and bend the metal with vise grips. this breaks the glue mount.
That then allows you to take the magnets off. They make great refrigerator magnets, or holding a book to a refrigerator as well.
802.11 interference issues are largely becuase it exists in an unlicensed band, thus no control over what is where.
802.16 can be done there also, but will actually be a mix of assigned spectrum (with license required) or some unlicensed also. Frequency allocations are still being worked out and will be assigned.
We never said there was a lot of wheat on this topic. Mostly chaff. The group is a lot of end users when it comes to RF and HW. Lots of knowledge here on SW and IT.
However, geeks come in many forms and flavors.
11M operators!!! giggle!!! That CB operator has a linear on board to get to himself up to a KW. We won't mention all the non-linear spectral splattering he is doing in the process!
"The slashdot collective has as much business talking about wireless networking as any room full of male gynecologists and cross dressers has talking about childbirth."
Ahem.... The slashdot crowd is mostly software and IT types. Mostly true. However there are a few of us out here who are hardware types, and even a few of us out here who do RF and wireless communication for a living.
If you want to talk IIP3, EIRP, multipath fading, NF, VSWR, and similar, some of us do speak that language. Otherwise. learn to seperate the wheat from the chaff as they say.
It is for the 2-11 GHz in both ISM and licensed bands as well. It is also multipath and non-line of sight in this spectrum. (this is 802.16A) Not perfectly over the hill stuff, but to a limited degree it will bounce off buildings and similar.
Higher bands (10-60 GHZ ? Not sure I do not have the standard open in front of me right now.) are just line of sight. Much higher data rates.
The Cisco system you allude to is point to point and may actually serve as the support path to a 802.16 downlink basestation.
That 802.16 basestation then serves a pile of subscriber stations. Point to Multi-Point
This is all old technology, it has been done before in a non-standard manner The key thing is that when you design to a standard, the compatibility is there, the cost goes through the floor, and the volumes (hopefully!) go through the roof.
It is about point to multipoint communication and about last mile access. There are provisions ongoing to give it limited mobility, but the emphasis is not cellular handoff friendly.
This is all about last mile access. Fibre to the house is too pricey, DSL and cable modems are limited and not always available.
Also, the target is not to replace 802.11, which is all about "in the house" kinds of ranges. Rather, an appropriate parallel is the TV converter box in your house, plugs into the wall and provides outputs compatible with the TV. This is a subscriber station with an ethernet output. 802.16 gets to your house, 802.11 gets you around your house.
However there probably would be a memory buffer in the MP3 player. Keeping the disk drive going all the time would suck a lot of juice. So it probably functiosn like the present disk drive ipod, which spins up, dumps music to RAM, and then shuts down the drive and plays out from RAM.
Makes sense, the disk cache is optimized with memeory to go over an interface. Not needed for a dedicated streaming media device. Just the RAM asociated with the streaming output would get the job done.
Last weekend, I spent 2 hours talking with a friend, he is a supply Sargeant, USMC - He just got back from several months in Iraq and Kuwait - Mostly down near Kuwait managing supplies and food for all the troops.
Food, water, medicine, electricity, stable government yes...
But 90% of the people there wouldn't know a computer if they fell over it.
Sorry - Rotational latency dominates the access time for a disk drive. If a disk has been defragmented, everything is in a neat order, so the seek time doesn't matter. Just reading the FAT then getting to the first data cluster in the chain matters. Developed the silly things for 15 year, so been there done that... Berkley..... Oh well..
Sorry - disk drives don't work in a vacuum, the heads require air to lift them off the platter.
BTW - IBM developed a few years back (1994-96)a disk drive with platters the size of a quarter.
(about the same size as a 1 Euro coin for you on the other side of the pond.) Consequently this is very old news.
The 48G is a full blown graphing machine.
I mean something simple like an HP32, HP42, HP11, HP15, HP16 -it is a matter of taste, when I need graphing and fancy stuff I go to something like MATLAB - but if the 48G works for you, fine with me.
regards -
After you use an RPN calculator for a while it becomes second nature. Very simple to use and very fast to use. No thought required.
The argument for the uninitiated is that it does use fewer keystrokes. Back in the 70's the calculator war was between HP and TI. The engineers largely went to RPN calculators.
HP no longer produces the wide range of RPN calculators that they used to, and consequently if you are looking for a simple RPN calculator, you end up buying them on Ebay at 200-500 dollars a pop.
Getting the current into the array is probably not an issue, there are plenty of CMP metal processes that can provide the multi-layer interconnect paths. I have designed IC's with 1-3 amps of current on them as a matter of course.
The thermal issue seems much more of an issue than the actual currents, also the high current array drivers, and the ballasting of diodes for current balance. A thermal nightmare. The big commercial displays are already a sea of coolong fans to keep them alive.
However, the cost factor is probably dominant.
Right now 12 inch wafers are cutting edge. Those wafers are multi thousand dollar exercises.
When was the last time you used a 12 inch monitor?
Size Matters!;-)
Also, if you lose one diode in the array, you end up trashing the whole thing. Yield loss would be huge on the wafers. (Remember when Intel was giving away Pentium uP's on a keychain? Thise were junk chips that had defects. Something like 50% yield loss where half the uP's were junk and got turned into keychains.)
What if the whole wafer has to be 100% good? Maybe one in every 100 wafers would be good for a display.
Capacitors leak.... i.e. slowly discharge over time. That said, no matter how much charge you put on a cap, when you come back in two weeks, the silly thing is going to be at zero volts.
Batteries can sit for months. Yes, they too go stale, but a capacitor goes flat a lot sooner.
He who truly solves the battery problem will die a very rich man.... (or woman...)
Minor detail - look at the wire going off "somewhere else" - Present battery technology (even the best Li-Ion/Polymer/Something stuff out there.) will not support the power needed to transmit from a cell phone.
A GSM cell phone sucks an amp of current at 35dbm (max power) output. It is impressive that the present battery sizes can support this now. Maybe when fuel cells replace batteries? Or maybe if people don't mind only 2 minutes of "talk" time before they have to recharge?
Smaller is not always better...
Slashdot Mirror
I have zero respect for patents....
Patents get issued all the time for nonsense, and things that do not work. A patent that I got, we proved (after filing) that it did not funtion as described. Two years later the patent still issued. Go figure...
If this gets challenged in court it will fall apart. Too much prior art. I would start with Morse Code...
After reading the information here (and the limited FCC stuff...) I believe that the term "WiFi" is being used incorrectly here.
Allocation of spectrum will lead to use, but the 802.11 protocalls are not optimized to this type of distance broadcasting on allocated spectrum.
My guess would be that 802.16a (a.k.a. "WiMax") might get used here, although the original frequency definitions for this were from 2-11GHz.
With the allocation of spectrum down under 700MHz, the RF front end definition of 802.16a, could be modified to go down to these bands without a lot of pain.
WiMax makes more sense here. Orderly calling between subscriber station and server tower, rather than the "bumper cars" mentality of WiFi. It is optimized to a multi-user environment.
WiFi was never designed for long distance communication or heavy multi-user environments. It amazes me that people attempt to use it for such. But then half the fun is seeing what you can do with it, rather than what it was actually designed for.
At 2GHz and up the QFN package is somewhat of a got to have, otherwise the inductance of the package leads starts to become a big issue.
That 14 pin DIP just won't hack it due to the lead frame inductance.
News for nerds.....
Obviously not communications nerds....
A MAC layer in communications terminology is the digital protocalls involved with all the control functions, ergo, priority, identity, and similar.
A PHY layer is the "physical layer" that is the RF frequencies involved, the antenna, the RF bandwidth of the channel and similar.
802.15.4 is designed for really low duty cycle use. Turns on once in a while, does its thing briefly, and then turns off. Think thermostats without a wire connection, or a set of burgalar alarm sensors without wires back to the alarm system.
Because of the low duty cycle it lends itself to battery power pretty well.
The Zigbee alliance is here:
http://www.zigbee.org
And the IEEE standards committee behind this effort is here:
http://www.ieee802.org/15/pub/TG4.html
Actually getting the MAC and PHY layers (now that you know what they are...) onto a single chip is pretty good thing. Most groups split it into a big digital ASIC (the MAC layer) and then an RF front end chip (the PHY layer)
Suggested option, put the metal that the magnets are glued to in a vise, and bend the metal with vise grips. this breaks the glue mount. That then allows you to take the magnets off. They make great refrigerator magnets, or holding a book to a refrigerator as well.
The 802.16A (WiMax) is going to do exactly this type of stuff.
A NAND gate takes 4 transistors in CMOS, 2 transistors in RTL....
802.11 interference issues are largely becuase it exists in an unlicensed band, thus no control over what is where.
802.16 can be done there also, but will actually be a mix of assigned spectrum (with license required) or some unlicensed also. Frequency allocations are still being worked out and will be assigned.
We never said there was a lot of wheat on this topic. Mostly chaff. The group is a lot of end users when it comes to RF and HW. Lots of knowledge here on SW and IT.
However, geeks come in many forms and flavors.
11M operators!!! giggle!!! That CB operator has a linear on board to get to himself up to a KW. We won't mention all the non-linear spectral splattering he is doing in the process!
vy 73
"The slashdot collective has as much business talking about wireless networking as any room full of male gynecologists and cross dressers has talking about childbirth."
Ahem.... The slashdot crowd is mostly software and IT types. Mostly true. However there are a few of us out here who are hardware types, and even a few of us out here who do RF and wireless communication for a living.
If you want to talk IIP3, EIRP, multipath fading, NF, VSWR, and similar, some of us do speak that language. Otherwise. learn to seperate the wheat from the chaff as they say.
802.16 is not just for the ISM band.
It is for the 2-11 GHz in both ISM and licensed bands as well. It is also multipath and non-line of sight in this spectrum. (this is 802.16A) Not perfectly over the hill stuff, but to a limited degree it will bounce off buildings and similar.
Higher bands (10-60 GHZ ? Not sure I do not have the standard open in front of me right now.) are just line of sight. Much higher data rates.
The Cisco system you allude to is point to point and may actually serve as the support path to a 802.16 downlink basestation.
That 802.16 basestation then serves a pile of subscriber stations. Point to Multi-Point
This is all old technology, it has been done before in a non-standard manner The key thing is that when you design to a standard, the compatibility is there, the cost goes through the floor, and the volumes (hopefully!) go through the roof.
802.16 is not about mobile communication.
It is about point to multipoint communication and about last mile access. There are provisions ongoing to give it limited mobility, but the emphasis is not cellular handoff friendly.
This is all about last mile access. Fibre to the house is too pricey, DSL and cable modems are limited and not always available.
Also, the target is not to replace 802.11, which is all about "in the house" kinds of ranges. Rather, an appropriate parallel is the TV converter box in your house, plugs into the wall and provides outputs compatible with the TV. This is a subscriber station with an ethernet output. 802.16 gets to your house, 802.11 gets you around your house.
There is not a memeory cache in the disk drive.
However there probably would be a memory buffer in the MP3 player. Keeping the disk drive going all the time would suck a lot of juice. So it probably functiosn like the present disk drive ipod, which spins up, dumps music to RAM, and then shuts down the drive and plays out from RAM.
Makes sense, the disk cache is optimized with memeory to go over an interface. Not needed for a dedicated streaming media device. Just the RAM asociated with the streaming output would get the job done.
Last weekend, I spent 2 hours talking with a friend, he is a supply Sargeant, USMC - He just got back from several months in Iraq and Kuwait - Mostly down near Kuwait managing supplies and food for all the troops.
Food, water, medicine, electricity, stable government yes...
But 90% of the people there wouldn't know a computer if they fell over it.
Sorry - Rotational latency dominates the access time for a disk drive. If a disk has been defragmented, everything is in a neat order, so the seek time doesn't matter. Just reading the FAT then getting to the first data cluster in the chain matters. Developed the silly things for 15 year, so been there done that... Berkley..... Oh well..
Sorry - disk drives don't work in a vacuum, the heads require air to lift them off the platter.
BTW - IBM developed a few years back (1994-96)a disk drive with platters the size of a quarter. (about the same size as a 1 Euro coin for you on the other side of the pond.) Consequently this is very old news.
The 48G is a full blown graphing machine. I mean something simple like an HP32, HP42, HP11, HP15, HP16 -it is a matter of taste, when I need graphing and fancy stuff I go to something like MATLAB - but if the 48G works for you, fine with me. regards -
After you use an RPN calculator for a while it becomes second nature. Very simple to use and very fast to use. No thought required.
The argument for the uninitiated is that it does use fewer keystrokes. Back in the 70's the calculator war was between HP and TI. The engineers largely went to RPN calculators.
HP no longer produces the wide range of RPN calculators that they used to, and consequently if you are looking for a simple RPN calculator, you end up buying them on Ebay at 200-500 dollars a pop.
Take and run an ethernet cable upo into the attic of the house.
Get above the brick walls and plop your 802.11 box up there. That way it will cover inside (straight down below)and also outside.
More power hacks?
Sigh... Anybody rember people running 2KW on the CB band?
Getting the current into the array is probably not an issue, there are plenty of CMP metal processes that can provide the multi-layer interconnect paths. I have designed IC's with 1-3 amps of current on them as a matter of course. The thermal issue seems much more of an issue than the actual currents, also the high current array drivers, and the ballasting of diodes for current balance. A thermal nightmare. The big commercial displays are already a sea of coolong fans to keep them alive. However, the cost factor is probably dominant. Right now 12 inch wafers are cutting edge. Those wafers are multi thousand dollar exercises. When was the last time you used a 12 inch monitor? Size Matters! ;-)
Also, if you lose one diode in the array, you end up trashing the whole thing. Yield loss would be huge on the wafers. (Remember when Intel was giving away Pentium uP's on a keychain? Thise were junk chips that had defects. Something like 50% yield loss where half the uP's were junk and got turned into keychains.)
What if the whole wafer has to be 100% good? Maybe one in every 100 wafers would be good for a display.
Capacitors leak.... i.e. slowly discharge over time. That said, no matter how much charge you put on a cap, when you come back in two weeks, the silly thing is going to be at zero volts. Batteries can sit for months. Yes, they too go stale, but a capacitor goes flat a lot sooner. He who truly solves the battery problem will die a very rich man.... (or woman...)
Minor detail - look at the wire going off "somewhere else" - Present battery technology (even the best Li-Ion/Polymer/Something stuff out there.) will not support the power needed to transmit from a cell phone. A GSM cell phone sucks an amp of current at 35dbm (max power) output. It is impressive that the present battery sizes can support this now. Maybe when fuel cells replace batteries? Or maybe if people don't mind only 2 minutes of "talk" time before they have to recharge? Smaller is not always better...