And I find it humorous that you are responding to my humorous attempt to extend your humor as though it were an attack on your manhood.
You made a joke, I made a joke. You misunderstood my joke, I explained it.
Ringworld was built by tnunctip, wasn't it.
on
Ask Larry Niven
·
· Score: 4, Interesting
You stated in Ringworld Engineers that Louis Wu was incorrect in assuming Ringworld was build by the Pac. My theory is that it was built by the tnuctipun, who also created the Pac. Am I correct?
Here's my reasoning: The Tree Of Live virus is just too convenient to have evolved naturally. Somebody made it - who? The tnuctipun. They wanted a race of warriors smart enough to use Soft Weapon level tech, but fanatical enough to resist the Thrint Power.
Now, being nicely paranoid the tnuctip would be unlikely to make a warrior race without some form of control. What better control than making them pathetically stupid and weak until a trigger event you can control, and after triggering them keeping them addicted to something you control, like Tree Of Life root. So the tnuctip could have worlds full of stupid monkeys, and when needed dust the worlds with Tree Of Life virus and BANG - instant army.
I'd also bet that ANY tnuctip "smells right" to any Pac.
Now, where did the tnuctip survive The Great Suicide command? In statis, of course. However, once out of statis they would need a safe place to be - a place shielded from the Pac, hard to get to, defensible from long range. I'll bet scrith blocks The Power. And by the time a Thrint could get over the edge of the Ringworld and start ordering folks around, "Hey, what's that violet glow around everythZZZZAP".
So the Pac probably found monkeys on Earth when they got here, but just "displaced" them.
I believe what killed packet (or, if you prefer, what caused packet not to take off more than it has) is the fact that most hams confuse buad and bits per second.
The FCC part 97 spec allows for 19.2 kBaud on 2 meters and 56 kBaud on 440MHz. The problem with using 440MHz is propagation - 440 doesn't go as far under normal conditions as 2 meters, so as a result if you want a wide area network you tend to want to use 2 meters so as to have fewer nodes.
Now, back in the days when I was active in packet, telephone modems were anywhere from 2400 bit per second to 9600 bit per second, and packet was 1200 bit per second. With overhead, collisions, and the fact that packet is a simplex system, you had an effective throughput closer to 300 bits per second. However, that wasn't ENORMOUSLY slower than landline modems, and the fact that you weren't tying up your phone line would offset that.
However, with the average telephone modem today delivering between 22kBit/sec to 56kBit/sec, and with broadband delivering between 128 kBit/sec to 3MBit/sec, the gulf between packet and modems is just too large to use packet for anything other than APRS, short text messaging, and other very small files like that.
However, what if you could have a 2 meter packet system that moved a peak of, say, 38.4kBit/sec, so that with overhead+collisions+simplex gave you 9600 bit/sec real throughput. Now, that is within spitting distance of phone modems.
"But you cannot do 38.4 kBit/sec on 2 meters - the FCC says so!" Wrong - I cannot do 38.4 kBaud on 2 meters.
But if I use C4FM (4 level FM signaling), I have 2 bits per symbol. A Baud is a symbol per second. At 9600 Baud, and 2 bits/symbol, that gives me 19.2 kBit/sec. Go to 16QAM, and I have 4 bits per symbol, or 38.4 kBit/sec at 9600Baud.
Given the large amount of spam I get originating from various DSL providers, I would conjecture this sort of thing is pretty common.
The spammers send out a <voice type="jethro"> U 2 kin make $$$$ width you're computer!!!!!! </voice> spam, and get multiple morons who figger (sic) "Gorsh! I kin pays fur my DeeEssEll with dis!" and run the spamware.
You, sir, are the only person making this point in a civil fashion.
Let me ask you this, then: Is the allegedly 64 bit version of WinXP truly 64 bits through and through, or is it like WinNT/Alpha, where the programs are still limited to a 2G virtual address space?
In other words, can you indeed mmap() a 10G file into memory?
Under WinNT/Alpha, you couldn't. You could pull some tricks IF AND ONLY IF you had enough physical memory to back the entire file (i.e. you could mmap() a 10G file IF you had 10G of physical memory), but you couldn't mmap() a file larger than your physical memory and let the pager do its thing.
Larry Niven (through the character of Louis Wu), Ringworld Engineers
I think that the main reason we see so much "grief counseling" these days is that people are not being encouraged to face their grief, accept it, and move on. Rather, they are encouraged to wallow in it like pigs in a mud hole - going over and over the problem, obsessing over it: "Oh, you still cannot face losing your favorite wallpaper image? Take another week off work - it'll be OK one day..." In a sick way, grieving makes the person empowered.
Yes, people feel grief. But there SHOULD be levels of grief - you might feel bad if your car dies, you should feel worse if your pet dies, and you should feel REALLY BAD if your mother dies.
But if you grieve for days because your car died, you need a counselor who will help you realize your grief is inappropriate.
Getting all bent out of shape over losing your data is just WRONG.
Please do the community a favor. Tomorrow, stand behind your husband with a large foam bat, and every time he misspells a word, or uses a non-word like "tho", club him.
I gottTHWACK ask thoTHWACK, at what point would it have made more sense just to buy a regular computer? The green lighting is a nice touch thoTHWACK.
On a project I designed, I deliberately designed the system to have TCL built-in, for a very simple reason.
Scripting has its place, as does more conventional compiled code.
Use compiled code to do the heavy lifting - in my case, things like FFTs, signal analysis, and such.
Use scripting to tie it all together.
That way, when you are trying to figure out the problem domain ("Now, what does the radio expect me to do when it sends a GTC message - maybe it wants a CASSN message? Clicky-click - No, doesn't seem to be it. Maybe a IDN message? Yep - that's it.") you can try things out very quickly.
You can also very quickly string together smaller functions into larger blocks ("Ok, to test the radio, first I do this, then that, then the other.")
I cannot even begin to imagine how long simple things would take if we didn't have an embedded scripting language.
I've wondered why more ISPs don't adopt this strategy to make themselves unattractive to spammers - so maybe you can tell me why it wouldn't work:
ISP has in their contract the following items: 1) a definition of spam, spamming, and spamming services. 2) A clause similar to the following: "The customer agrees not to spam, not to advertise any services hosted by the ISP via spam, not to provide services to promote spam. In the event of a violation, the customer will forfit US$10,000 clean up fees."
You require the customer to either a) put up the money in a bond, or b) put up a credit card.
Should the customer spam, and then try to back out on the credit card (dispute the charges), then you nail them with felony fraud charges, as they obviously never intended upon paying the bill in the first place.
The Intel answer allows for a chip to have more than 4G of physical memory in much the same way the old LIM EMS boards allowed a 8086 to have more than 1M of memory - it is a form of bank switching.
True, you could have a PIII with 10G of memory on it (in theory, anyway), but this would not help you for the common applications for which you need these quantities of memory - databases, video editing and so on.
In those tasks, you have ONE program that needs lots of memory. You ideally want to be able to take a multi-gigabyte file, and mmap() it so that it appears to your program to be just a stretch of memory. Then you can access the file with a simple pointer, and moving within the file is nothing more than pointer manipulation. You don't have to worry about paging the file in and out - that is the OS's virtual memory manager's problem.
PAE won't help you in those cases. At best, you can back some of the buffer cache with the PAE memory, creating in effect a glorified RAM disk.
PAE is great if you have a machine running hundreds of processes, each of which takes 100M of space. But this usually is NOT the case.
Just as machines with more than 1M of memory started out the providence of the high-end user and slowly moved down, 64 bit address space on the desktop will start out the providence of the high-end folks first, then will move down as it becomes more common.
I would guess the likely sequence will be something like:
1) We *nix folks had it first - I was running 64 bits on my Alpha years ago. But we are not "the masses", and so will be ignored by the mainstream. 2) The Macs will be next - Apple will port MacOS X to the newer 64 bit Power chips. This will greatly simplify video editing - one of Apples favorite areas to compete in. 64 bit Apple will make the Mac the chosen platform for video editing of large files (NOTE: a 40 minute capture from my Firewire camcorder is a couple of gig - so already the home consumer is getting close to needing this.) 3) Windows will finally release a 64 bit OS (also note: they could have done this YEARS ago under Alpha, but didn't - Windows NT under Alpha only could access a 32 bit address space.) Microsoft will hail this as a revolutionary breakthrough - "Windows AYCABTU is the first 64 bit OS for the home user!" *nix and Apple users will scratch their heads in puzzlement.
Well, I meant that 100 MSample/sec flash converter chips are available, not so much that they are available on cards.
However, I have seen "'scope on a card" cards that would do over a gigasample per second for a short period of time. Obviously you don't stream that kind of bandwidth to across PCI.
AMPS is the Advanced Moble Phone System - the system commonly in use in the USA (it was named a long time ago - and at that time it WAS advanced).
As for DoS'ing a cell tower - yes and no. A CDMA signal is much less affected by a narrowband carrier than GSM would be due to it being spread spectrum - the system will just work around the interference.
GSM used TDMA (time domain multple access) - each phone transmits only for a short period of time, at an assigned time. There are six time slots, and each phone gets one or two of those time slots.
You might therefore wipe out up to six conversations by jamming one channel in GSM.
However, the cell sites will immediately note the interference, and will move the calls off that frequency. And the site will be able to determine which sector the interference is in (a sector is 60 degrees). Multiple sites will report, and so they will have your position pretty quickly. Do this for any length of time and the black vans will roll. And they WILL track your sorry ass down very quickly, and you WILL be looking at striped sunlight.
However I reckon the reason that the card costs $1300 instead of a mere $200 is because the company is trying to recoup R&D costs as well as turn a tidy profit.
Yes and no. This card is a standard card that folks like me would use to build a system, so they don't sell in quantity, so the non-recurring engineering costs aren't amortized over a large number of cards.
However, this card is also pretty over-engineered for what the GnuRadio folks use - that is why I think that a dedicated RF capture card would a) do a better job and b) be less expensive overall.
(now, I wait for the 2 minute comment timeout. Dope-de-doe).
Well, when I do the IF filter designs for the digital sections, I generally use Matlab/Signal Processing tool or Mathcad - if you think I to the inverse FFTs by hand you are crazy. But yes, when I was an undergrad, I had to work through all the equations by hand - an assignment took hours to complete</voice>.
However - yes, if nothing else start playing around with DSP using your soundcard - practical experience rules!
As for your classmates - yeah, I had fun during my labs: them: "OK, which one is the 1k resistor?" Me: "Uhh... This one." them: "How do you know? You didn't measure it!" Me: "Brown, Black, Red - 1K" Them: "Huh?"
Of course, I was the guy who brought his own logic probe, ICs, prototyping board, oscilloscope, and databook to digital logic class - my brother had given me these back when I was 12 (Thank you, Bob - RIP).
And I do almost ALL my work in C++ - without multiple inheritance Java won't cut it (and speed is king, as well).
I agree with BigBlockMopar - I regret not taking the extra time to get my BAMath - if you get a BSEE you are only a few credit hours from a BAMath. And for this sort of work, LaPlace is your buddy - almost all this work involves working in the LaPlace domain (or transforming into the z domain for digitial systems.)
I'm rather the inverse of BigBlockMopar - I don't do the electromagnetic field theory, rather I do the modulation and software engineering work.
Get a degree, get a ham license, get into a shop that does this - Motorola, Ericcsson, E.F. Johnson, Agilent, IFR/Aeroflex, anybody like that.
We don't make the board as a seperate product - it is a part of a piece of test equipment we sell. What makes our equipment worth the money is the total package - software and hardware. GnuRadio might not be a competitor, but.... Once you can digitally recover the signal, doing the parametrics on it is just more programming. So in theory GnuRadio could be a competitor.
<personal opinion !employer opinion> I'd welcome the chance to work with GnuRadio - most of the protocols we work with are patented and licensed, and GnuRadio probably couldn't implement them legally, and so couldn't compete. However, I'm not a C?O - just a senior engineer. I don't set policy, and somehow I doubt the muckety-mucks would go for this. </personal opinion>
Actually, all I have is a BSEE. That, and 13 years experience in this field.
And what, exactly, are you wanting to decode? The stuff between 70 MHz and 130 MHz you could probably do with the simpler solution that GnuRadio targets - basically a receiver and your sound card.
Narrowband stuff like broadcast FM and the like are pretty easy to do - it's wideband stuff like video that gets hard.
I don't know - it could be that their receiver section has a lot of phase noise, or isn't very linear, or is trying to AGC the signal, or isn't quite wide enough, or isn't very flat through the IF passband.
They may also not have the strongest signal to recover, and are getting a high bit error rate.
They may also have a bug in their forward error correction routines (or not have ANY FEC).
They may not be feeding an optimal level into their A/D converter, so they are not using the full range.
I'm not sure what you mean by "donating the hardware" - do you mean giving the GnuRadio folks a sample board, or what?
The problem is this: even if I wanted to give them a sample board, I could very will a) get in trouble with my employer, since that's what we do for a living, and b) get the GnuRadio folks in trouble by causing them to infringe on my employer's intellectual property.
That's the rub - I'm too close to this thing to safely get seriously involved. It would be like one of the core folks on the Windows software team being involved in the Wine project.
6 Msample/sec at 12 bits is a little weak for this sort of thing - I'd either want more bits or more speed - prefereably more speed, like 40 MSample/sec or better.
The 12 bits in the RX is not really the limiting factor.
A 12 bit A/D raises your quantization noise, but the quantization noise is at the sampling frequency - you can "shape" the noise away from the signal of interest, and gain bits as you process the signal.
This is the trick that the 1 bit D/A sections in modern CD players use - the D/A is only 1 bit, but running at a very high rate (1MHz or more). As a result, the quantization noise is all above the audio band, and the system has the low-frequency resolution of a 16 bit system.
Think of it as dithering - really, that is what it is.
Slashdot Mirror
And I find it humorous that you are responding to my humorous attempt to extend your humor as though it were an attack on your manhood.
You made a joke, I made a joke. You misunderstood my joke, I explained it.
You stated in Ringworld Engineers that Louis Wu was incorrect in assuming Ringworld was build by the Pac. My theory is that it was built by the tnuctipun, who also created the Pac. Am I correct?
Here's my reasoning:
The Tree Of Live virus is just too convenient to have evolved naturally. Somebody made it - who? The tnuctipun. They wanted a race of warriors smart enough to use Soft Weapon level tech, but fanatical enough to resist the Thrint Power.
Now, being nicely paranoid the tnuctip would be unlikely to make a warrior race without some form of control. What better control than making them pathetically stupid and weak until a trigger event you can control, and after triggering them keeping them addicted to something you control, like Tree Of Life root. So the tnuctip could have worlds full of stupid monkeys, and when needed dust the worlds with Tree Of Life virus and BANG - instant army.
I'd also bet that ANY tnuctip "smells right" to any Pac.
Now, where did the tnuctip survive The Great Suicide command? In statis, of course. However, once out of statis they would need a safe place to be - a place shielded from the Pac, hard to get to, defensible from long range. I'll bet scrith blocks The Power. And by the time a Thrint could get over the edge of the Ringworld and start ordering folks around, "Hey, what's that violet glow around everythZZZZAP".
So the Pac probably found monkeys on Earth when they got here, but just "displaced" them.
Am I even remotely correct?
What I meant was that only quarks can "be" down, as in the down quark.
An electron may have a down spin, but only a quark can "be" down.
I believe what killed packet (or, if you prefer, what caused packet not to take off more than it has) is the fact that most hams confuse buad and bits per second.
The FCC part 97 spec allows for 19.2 kBaud on 2 meters and 56 kBaud on 440MHz. The problem with using 440MHz is propagation - 440 doesn't go as far under normal conditions as 2 meters, so as a result if you want a wide area network you tend to want to use 2 meters so as to have fewer nodes.
Now, back in the days when I was active in packet, telephone modems were anywhere from 2400 bit per second to 9600 bit per second, and packet was 1200 bit per second. With overhead, collisions, and the fact that packet is a simplex system, you had an effective throughput closer to 300 bits per second. However, that wasn't ENORMOUSLY slower than landline modems, and the fact that you weren't tying up your phone line would offset that.
However, with the average telephone modem today delivering between 22kBit/sec to 56kBit/sec, and with broadband delivering between 128 kBit/sec to 3MBit/sec, the gulf between packet and modems is just too large to use packet for anything other than APRS, short text messaging, and other very small files like that.
However, what if you could have a 2 meter packet system that moved a peak of, say, 38.4kBit/sec, so that with overhead+collisions+simplex gave you 9600 bit/sec real throughput. Now, that is within spitting distance of phone modems.
"But you cannot do 38.4 kBit/sec on 2 meters - the FCC says so!" Wrong - I cannot do 38.4 kBaud on 2 meters.
But if I use C4FM (4 level FM signaling), I have 2 bits per symbol. A Baud is a symbol per second. At 9600 Baud, and 2 bits/symbol, that gives me 19.2 kBit/sec. Go to 16QAM, and I have 4 bits per symbol, or 38.4 kBit/sec at 9600Baud.
How can an electron be down? Quarks can be down, but leptons cannot be.
The first particle isn't destroyed in quantum teleportation. What is destroyed is the quantum state of they particle - what spin/polarization it had.
To put it very loosely - in quantum teleportation, the original object would be scrambled. But it would still exist as mass.
Given the large amount of spam I get originating from various DSL providers, I would conjecture this sort of thing is pretty common.
The spammers send out a
<voice type="jethro">
U 2 kin make $$$$ width you're computer!!!!!!
</voice> spam, and get multiple morons who figger (sic) "Gorsh! I kin pays fur my DeeEssEll with dis!" and run the spamware.
You, sir, are the only person making this point in a civil fashion.
Let me ask you this, then: Is the allegedly 64 bit version of WinXP truly 64 bits through and through, or is it like WinNT/Alpha, where the programs are still limited to a 2G virtual address space?
In other words, can you indeed mmap() a 10G file into memory?
Under WinNT/Alpha, you couldn't. You could pull some tricks IF AND ONLY IF you had enough physical memory to back the entire file (i.e. you could mmap() a 10G file IF you had 10G of physical memory), but you couldn't mmap() a file larger than your physical memory and let the pager do its thing.
I think that the main reason we see so much "grief counseling" these days is that people are not being encouraged to face their grief, accept it, and move on. Rather, they are encouraged to wallow in it like pigs in a mud hole - going over and over the problem, obsessing over it: "Oh, you still cannot face losing your favorite wallpaper image? Take another week off work - it'll be OK one day..." In a sick way, grieving makes the person empowered.
Yes, people feel grief. But there SHOULD be levels of grief - you might feel bad if your car dies, you should feel worse if your pet dies, and you should feel REALLY BAD if your mother dies.
But if you grieve for days because your car died, you need a counselor who will help you realize your grief is inappropriate.
Getting all bent out of shape over losing your data is just WRONG.
Dear Mrs. Malda,
Please do the community a favor. Tomorrow, stand behind your husband with a large foam bat, and every time he misspells a word, or uses a non-word like "tho", club him.
I gottTHWACK ask thoTHWACK, at what point would it have made more sense just to buy a regular computer? The green lighting is a nice touch thoTHWACK.
On a project I designed, I deliberately designed the system to have TCL built-in, for a very simple reason.
Scripting has its place, as does more conventional compiled code.
Use compiled code to do the heavy lifting - in my case, things like FFTs, signal analysis, and such.
Use scripting to tie it all together.
That way, when you are trying to figure out the problem domain ("Now, what does the radio expect me to do when it sends a GTC message - maybe it wants a CASSN message? Clicky-click - No, doesn't seem to be it. Maybe a IDN message? Yep - that's it.") you can try things out very quickly.
You can also very quickly string together smaller functions into larger blocks ("Ok, to test the radio, first I do this, then that, then the other.")
I cannot even begin to imagine how long simple things would take if we didn't have an embedded scripting language.
I've wondered why more ISPs don't adopt this strategy to make themselves unattractive to spammers - so maybe you can tell me why it wouldn't work:
ISP has in their contract the following items:
1) a definition of spam, spamming, and spamming services.
2) A clause similar to the following: "The customer agrees not to spam, not to advertise any services hosted by the ISP via spam, not to provide services to promote spam. In the event of a violation, the customer will forfit US$10,000 clean up fees."
You require the customer to either a) put up the money in a bond, or b) put up a credit card.
Should the customer spam, and then try to back out on the credit card (dispute the charges), then you nail them with felony fraud charges, as they obviously never intended upon paying the bill in the first place.
The Intel answer allows for a chip to have more than 4G of physical memory in much the same way the old LIM EMS boards allowed a 8086 to have more than 1M of memory - it is a form of bank switching.
True, you could have a PIII with 10G of memory on it (in theory, anyway), but this would not help you for the common applications for which you need these quantities of memory - databases, video editing and so on.
In those tasks, you have ONE program that needs lots of memory. You ideally want to be able to take a multi-gigabyte file, and mmap() it so that it appears to your program to be just a stretch of memory. Then you can access the file with a simple pointer, and moving within the file is nothing more than pointer manipulation. You don't have to worry about paging the file in and out - that is the OS's virtual memory manager's problem.
PAE won't help you in those cases. At best, you can back some of the buffer cache with the PAE memory, creating in effect a glorified RAM disk.
PAE is great if you have a machine running hundreds of processes, each of which takes 100M of space. But this usually is NOT the case.
Just as machines with more than 1M of memory started out the providence of the high-end user and slowly moved down, 64 bit address space on the desktop will start out the providence of the high-end folks first, then will move down as it becomes more common.
I would guess the likely sequence will be something like:
1) We *nix folks had it first - I was running 64 bits on my Alpha years ago. But we are not "the masses", and so will be ignored by the mainstream.
2) The Macs will be next - Apple will port MacOS X to the newer 64 bit Power chips. This will greatly simplify video editing - one of Apples favorite areas to compete in. 64 bit Apple will make the Mac the chosen platform for video editing of large files (NOTE: a 40 minute capture from my Firewire camcorder is a couple of gig - so already the home consumer is getting close to needing this.)
3) Windows will finally release a 64 bit OS (also note: they could have done this YEARS ago under Alpha, but didn't - Windows NT under Alpha only could access a 32 bit address space.) Microsoft will hail this as a revolutionary breakthrough - "Windows AYCABTU is the first 64 bit OS for the home user!" *nix and Apple users will scratch their heads in puzzlement.
Well, I meant that 100 MSample/sec flash converter chips are available, not so much that they are available on cards.
However, I have seen "'scope on a card" cards that would do over a gigasample per second for a short period of time. Obviously you don't stream that kind of bandwidth to across PCI.
OK, perhaps you could tell me WHAT you are wanting to listen to - then I could probably tell you the protocol.
But if you are looking at a 2 MBaud signal, you aren't going to be able to do this with simple stuff - you will need a fast digitizer card.
And if it is a QPSK or QAM signal, you are going to need a good tuner.
AMPS is the Advanced Moble Phone System - the system commonly in use in the USA (it was named a long time ago - and at that time it WAS advanced).
As for DoS'ing a cell tower - yes and no. A CDMA signal is much less affected by a narrowband carrier than GSM would be due to it being spread spectrum - the system will just work around the interference.
GSM used TDMA (time domain multple access) - each phone transmits only for a short period of time, at an assigned time. There are six time slots, and each phone gets one or two of those time slots.
You might therefore wipe out up to six conversations by jamming one channel in GSM.
However, the cell sites will immediately note the interference, and will move the calls off that frequency. And the site will be able to determine which sector the interference is in (a sector is 60 degrees). Multiple sites will report, and so they will have your position pretty quickly. Do this for any length of time and the black vans will roll. And they WILL track your sorry ass down very quickly, and you WILL be looking at striped sunlight.
Yes and no. This card is a standard card that folks like me would use to build a system, so they don't sell in quantity, so the non-recurring engineering costs aren't amortized over a large number of cards.
However, this card is also pretty over-engineered for what the GnuRadio folks use - that is why I think that a dedicated RF capture card would a) do a better job and b) be less expensive overall.
(now, I wait for the 2 minute comment timeout. Dope-de-doe).
Well, when I do the IF filter designs for the digital sections, I generally use Matlab/Signal Processing tool or Mathcad - if you think I to the inverse FFTs by hand you are crazy. But yes, when I was an undergrad, I had to work through all the equations by hand - an assignment took hours to complete</voice>.
However - yes, if nothing else start playing around with DSP using your soundcard - practical experience rules!
As for your classmates - yeah, I had fun during my labs:
them: "OK, which one is the 1k resistor?"
Me: "Uhh... This one."
them: "How do you know? You didn't measure it!"
Me: "Brown, Black, Red - 1K"
Them: "Huh?"
Of course, I was the guy who brought his own logic probe, ICs, prototyping board, oscilloscope, and databook to digital logic class - my brother had given me these back when I was 12 (Thank you, Bob - RIP).
And I do almost ALL my work in C++ - without multiple inheritance Java won't cut it (and speed is king, as well).
I agree with BigBlockMopar - I regret not taking the extra time to get my BAMath - if you get a BSEE you are only a few credit hours from a BAMath. And for this sort of work, LaPlace is your buddy - almost all this work involves working in the LaPlace domain (or transforming into the z domain for digitial systems.)
I'm rather the inverse of BigBlockMopar - I don't do the electromagnetic field theory, rather I do the modulation and software engineering work.
Get a degree, get a ham license, get into a shop that does this - Motorola, Ericcsson, E.F. Johnson, Agilent, IFR/Aeroflex, anybody like that.
We don't make the board as a seperate product - it is a part of a piece of test equipment we sell. What makes our equipment worth the money is the total package - software and hardware. GnuRadio might not be a competitor, but.... Once you can digitally recover the signal, doing the parametrics on it is just more programming. So in theory GnuRadio could be a competitor.
<personal opinion !employer opinion>
I'd welcome the chance to work with GnuRadio - most of the protocols we work with are patented and licensed, and GnuRadio probably couldn't implement them legally, and so couldn't compete. However, I'm not a C?O - just a senior engineer. I don't set policy, and somehow I doubt the muckety-mucks would go for this.
</personal opinion>
Actually, all I have is a BSEE. That, and 13 years experience in this field.
And what, exactly, are you wanting to decode? The stuff between 70 MHz and 130 MHz you could probably do with the simpler solution that GnuRadio targets - basically a receiver and your sound card.
Narrowband stuff like broadcast FM and the like are pretty easy to do - it's wideband stuff like video that gets hard.
I don't know - it could be that their receiver section has a lot of phase noise, or isn't very linear, or is trying to AGC the signal, or isn't quite wide enough, or isn't very flat through the IF passband.
They may also not have the strongest signal to recover, and are getting a high bit error rate.
They may also have a bug in their forward error correction routines (or not have ANY FEC).
They may not be feeding an optimal level into their A/D converter, so they are not using the full range.
I'm not sure what you mean by "donating the hardware" - do you mean giving the GnuRadio folks a sample board, or what?
The problem is this: even if I wanted to give them a sample board, I could very will a) get in trouble with my employer, since that's what we do for a living, and b) get the GnuRadio folks in trouble by causing them to infringe on my employer's intellectual property.
That's the rub - I'm too close to this thing to safely get seriously involved. It would be like one of the core folks on the Windows software team being involved in the Wine project.
6 Msample/sec at 12 bits is a little weak for this sort of thing - I'd either want more bits or more speed - prefereably more speed, like 40 MSample/sec or better.
The 12 bits in the RX is not really the limiting factor.
A 12 bit A/D raises your quantization noise, but the quantization noise is at the sampling frequency - you can "shape" the noise away from the signal of interest, and gain bits as you process the signal.
This is the trick that the 1 bit D/A sections in modern CD players use - the D/A is only 1 bit, but running at a very high rate (1MHz or more). As a result, the quantization noise is all above the audio band, and the system has the low-frequency resolution of a 16 bit system.
Think of it as dithering - really, that is what it is.