Slashdot Mirror
Software Defined Radio Systems
sundbug writes: "Very good article on CommVerge about a new technology called Software Defined Radio Systems. Pretty cool to have one computer receive and potentially send over several protocols."
← Back to Stories (view on slashdot.org)
Is it just my browser, or is there a monkey in the middle of the text on that web page?
-S
It may be economical in certain situations, but like FPGA designs you pay for the flexibility - if you don't need that flexibility it's a waste of money. So while basestations might usefully incorporate this technology, one can't help thinking that handsets - cheap and easy to replace - would not benefit so much.
You know, this would be wonderful. Now all the l33t h4x0rz don't need to buy a $200 radio scanner to know that the FBI is coming to take them down for cracking that zip file full of pr0n the night before. Think about it. You could tune in to your neighbor's cordless phone and listen to them battle it out. You could tune in to the chick down the street and listen to her steamy tales of PK on UO/EQ/AO. All in the privacy of your own home! But, then again, where would the limit be? If there were no frequency limit, you could listen to about anything! I suspect that the whole thing of ownership of radiowaves would come up (but their radio waves cross into MY property, I'm ENTITLED to listen in on the phonesex going on next door...). I suspect a move to phones that are corded, with a double-shielded cord. Or a phone that uses 256bit encryptation.
Job? I don't have time to get a job! Who will sit around and bitch about being broke and unemployed then?
They have been doing software based modulation in professional grade Radio Contol Aircraft Systems for a number of years now. If a particular frequency was being taken then select a frequency from a menu on your transmitter and instruct your receiver to do the same.
It has been a while since I followed the technology so I am sure that it has improved since I last saw it.
This would be a sweet toy if they could get it in say USB and for under $200. This would replace any type of scanner anyone would want to use, and let you have a lot of fun to. Unfortunatly, this will not happen. Unless someone hacks the unit and gets a nockoff out on the market.
Just think, Private Video broadcasts (like X-10), TV, Satelite broadcasts, Police, Military, aircraft, cell phones, coreless phones, FM radio, Racecar Cams, HDTV, WI-FI, and so on. All from one little box on your PC/laptop.
Who wants Pork Chops?
What a bunch of fluff. It pisses me off when an MBA walzes in & their genius revelation is that it would be nice if "everything worked together". Fundamental differences at all layers of the OSI model, combined with competing technologies that won't open up their systems make this vision unobtainable. Yeah, let's weave together this short range wireless token ring network with this IP net - all we have to do is NAT all boxes in the token ring net. Then we have cell phones - don't have any IPs for them, guess we want them to talk IPv4 since the rest already is - NAT them too. Shit, is the garage door opener supposed to be on this magical pixie network? Screw it, time for IPv6 - OK everyone, upgrade everything you have to talk IPv6, and you garage door opener manufacturers schedule a tech meeting with these cellular manufacturers over here because your frequencies overlap. What were we doing again? oh yeah, trying to get actual work done at the nodes instead of playing with the paths.
Nice suit. Get out of the engineering department. we have real work to do.
Good Christ! That's nearly as many times as Redhat has changed their direcory structure! :p
One thing that made Amateur Radio appealing to a lot of the old timers was the fact that they could build their own transmitters and receivers and antennas. With Software Defined Radio Systems, hams could build their own antennas and code their own radios.
Code could be shared among enthusiasts that worked particularly well for particular applications... Hard core hams could overclock their systems to run serious number crunching code for enhanced dynamic range to pull out weak signals, etc.
KF8QE
Amazing magic tricks
I have been wondering about technology like this for a while. More specifically, the possibility to build cellular stations that are compatable with every communications protocol. If a company could build small communication units that were completely reprogrammable, they could place them around cities and open lease agreements with various phone and communication companies.
I'm not familiar enough with FCC regulations to know that someone could actually build a business model around this, but the idea still intrigues me. We're going to be using radio based gadgets for the foreseeable future, and a company that could move itself to become an all purpose wireless provider would have a good deal of potential.
Motorola was working on a chipset for the cell phone market that was highly programmable - it could simulate differing standards on the fly. What killed the project was horrible battery drain - hopefull these people have figgured it out.
Moneyed corporations, non-working 'poor' and criminal prisoners are turning productive citizens into tax-slaves.
...is that when people start talking about everything playing well with other devices there is a problem.
It could even function as an FRS (family radio service) walkie-talkie, listen in on your baby monitor, and act as a garage-door opener.
I am of the opinion that if I can monitor my baby and open my garage door, so can someone else. And with the lack of security that most people tend to apply to their equipment I do not like this idea at all.
MessEdUp
#/var/www/v
Swap mp3's while driving down the freeway.
What could be better?
"Would it kill you to put down the toilet seat?" -- Maya Angelou
My machine already receives and sends over several protocols. TCP, HTTP, SSL, FTP, SMTP, POP, IMAP and many many *many* others.
People, Motorola has built this box a LONG
time ago. I guess the only reason why people
don't know about it is due to the fact that
it was purpose built for a rather nasty
government agency thats rather uptight with its
secrets...Basic idea is
1) download transform into PPC
2) start A/D
3) transform on PPC
4) pipe out on D/A
DUH...
The digital implementation sucks up the whole cellular band, downconverts it to a traditional IF with an oscillator and mixer, applies IF filters, and then digitizes the result. A DSP system then simulates all the receivers simultaneously.
(FFT the whole band once, then read out the value for each channel.) The trick is having enough compute power.
The
GNU Radio project is doing stuff like this. There's some pretty neat code already developed, using cheap A/D converters and varactor tuners from old TV sets. Cool stuff.
TI has been selling "programmable DSP's" for just this application for years. You build your tower and it supports the two most popular protocols of the day. Next year a new protocol comes out, rather than replacing the tower(or phone) you re-program the DSP. Hardly news...
If you build such base stations and sell them, there goes your hardware business. In my company (GSM cell phone base stations) we are JUST starting to sell software as a separate product..as in last year for the first time. Before that all sales were hardware.
why we all think software-defined radios are cool but software-defined modems are crap?
Well, I am not directly involved with SDR tech yet in ham radio, but I understand the tech, and when I have the time, I will get involved. Anyway, here is the link: http://www.tapr.org/sdr/
So, any hams that want to get involved, there is a start, I know I will be some time in the next 12 months, I'm really hanging out for a HF version, or maybe even a 6m one, maybe I should design one...
VK3TST
VK3TST
-- "People aren't stupid. Usually." -- jd
SDR can implement the old protocols but not the new ones.
It takes complicated signal structures and protocols to pack the most information into the least bandwidth. With bandwidth getting more expensive in the FCC auctions, wireless service providers require the most effective (and complex) protocols. Basically they implement what the technology supports after very careful study...$Billions ride on getting the choice right. The most cost effective designs use bleeding edge custom ASIC's.
The performance available from custom chips far exceeds what someone can implement in a DSP or general purpose processor (or even an FPGA).
Sure you can wait 5-10 years and software will catch up. But by then, the wireless operators will have upgraded to a new protocol that software can't do.
Did anyone else look at the html on that page? Holycrap it is loaded with a ton of CSS and JavaScript. I was trying to figure out why that page scrolled so slowly... There's the reason!
Large print giveth, and the small print taketh away
check out monkeysvsrobots.com. You'll have plenty to keep you happy. =)
Large print giveth, and the small print taketh away
And I agree with several of the posters - I'd like to see this sort of thing work its way into a box next to the computer.
Take
this bad boy, a four channel programmable down converter - 4 radios on a chip. You feed in 1 to 4 IF data streams, and this guy will decode them - about 2 billion operations per second, on a chip the size of your thumbnail (micro-ball-grid array). I work with its little brother, the 50214, on
my project, and I can't wait until I get past the big stuff and get some time to play.
That's the sick thing about soft radios: you do one down conversion from RF to IF, then digitize it, and from there on it's all math. When you are a ham operator, a math geek, and a software engineer, and you get paid to play with these, well, life is good.
www.eFax.com are spammers
wow, i didnt expect that one.(no one expects the spanish inquisition.)
-- free as in swatantryam - not soujanyam.
It's still very expensive to do a true SDR that can suck in DC to daylight (or any reasonable subset thereof) and digitize it. However, you can take beginning steps by using traditional methods for the "front end" of the radio, and using DSP techniques at the back end. Hams are doing that now, and we're seeing some very interesting results.
Bob Larkin, W7PUA, developed a DSP radio called the DSP-10 that works at the ham 2M (144-148MHz) band, with a DSP board that's essentially a souped-up sound card. In addition to handling normal communication modes, Bob programmed the DSP for several unique modes that involve extremely narrow bandwidths (easy for DSP, virtually impossible to do in analog) and very long data integration times.
The result is that a pair of these radios were able to talk to each other via "moonbounce" (yes, bouncing signals off the moon) with power levels and antennas far below the macho amplifiers and antennas normally required.
The DSP-10 software is fully open source, so it's wide open for experimenters to work with. The radio itself is being sold as a kit by TAPR (http://www.tapr.org), a ham radio R&D organization. Details on the DSP-10 are at http://www.tapr.org/tapr/html/Fdsp10.html.
This idea is not new. It has lots of good potential. There is only one tiny detail which most of you computer geeks seem to ignore:
Dynamic Range!
Commercial grade SW receivers often have dynamic ranges of more than 100 dB (depending on how you measure it, of course).
You won't find A/D converters with that dynamic range and bandwidth greater than audio at a price you would be willing to pay. Most high end receivers do use DSP, but only at a low, heavily filtered IF, and only after some AGC processing. It's more of a demodulation effort than anything else.
Frankly, "primitive" analog gear can still run rings around the performance you'd get from SDR. And don't get me started about what Direct Digital Synthysis LO's will do to your performance. Did most of you know that it's standard practice to use a frequency locked loop and a VCO to clean up the spectral garbage you get from DDS?
SDR is a nice development path. Some day, when the A/D technology is up to the task, it will be really cool stuff. But right now, I wouldn't expect much more performance from such a radio than you'd get from a wide-band scanner.
As for me, I'd rather build a clean VFO against a KK7B R2, and then digitize the I and Q channels for audio processing.
I happen to be an RF engineer, and part time linux-geek... My company designs both the radio (RF/IF) sections of mobile handsets as well as the kick-ass ADCs that would sample the entire IF spectrum of one of these software radios. I've seen a few posts here that tells me that people are thinking these SDRs can do more than they really can.
For the sake of simplicity, lets just talk about SDRs as they might be used for mobile phones (apposed to wireless LAN or HAM applications). Every handset I've ever seen is based on the familiar superheterodyne architecture, where you modulate your baseband on an IF carrier, and then upconvert this IF to RF.
When talking about a typical Motorola CDMA handset, that baseband is about 1250kHz wide (read up on direct sequence spread spectrum to understand why its so wide), and the data is dqpsk-modulated on a 220MHz or so carrier -- this section is referred to as the modulator... that IF is then mixed with the RF local oscillator to get upconverted to the correct RF channel -- this is called the upconverter block... from here, it goes to a linear power amplifier, out through a duplexer (to keep the TX from blowing out the receiver, which listens as the transmitter talks), and out through the antenna.
A "software radio" approach eliminates the modulator/demodulator block, and uses fast analog-to-digital converters (ADCs) to sample the IF, and fast DACs to generate it... lots of crazy DSP allows you to emulate/listen to whatever modulation scheme you want (as simple as dqpsk or as intense as 64QAM).
So yes, we now have the ability to change modulation schemes as SNR degrades, or incrementally rev your WDCMA work-in-progress, or move to a new vocoder -- but all of this flexibility is limited to the mod/demod and baseband blocks -- you _cannot_ change things like:
() the RF band
-> which means this does not allow you to easily have a handset that is a 800MHz CDMA phone one moment, a HAM transciever the next, and a piece of 40GHz spy equipment the next... in fact, you won't even be able to listen to your neighbors' 900MHz cordless phone conversations, because they're from 902-928MHz, and you're at 869-894MHz... likewise, by itself, this won't allow you to take your US GSM phone to Europe, because they're on a different band.
being able to change modulation schemes cannot retune a passive SAW filter, or make an oscillator run at a frequency it was designed to.
() the output power and linearity of the power amplifier
-> GSM is a non-linear (constant evelope) modulation scheme at 4W peak -- all info is in frequency modulation, none in amplitude modulation... CDMA uses both frequency (well, phase) and amplitude, and thus requires a linear PA, but less than 1W peak. the amplifier that would work well for both would be big and very inefficienct -- an SDR solution cannot get around this
() the IF frequency
-> these concerns go mostly to 'frequency planning' concerns that the radio system designer must deal with... different considerations for different systems (GSM, TDMA, CDMA, etc)
() lots of other limitations having to do with time-domain multiplexing, PA power control, RF filtering...
with regards to the ADCs and DACs required to implement these SDRs, as they stand now, the handset can't afford the power draw, which might be from 100mW to 400mW.
generally, SDRs are more suited to a basestation application, where size and power can be traded off for flexibility... even then, most of these limitations still apply.
i'm by no means an expert, but i try to pick their brain as much as i can... i'll keep track of this thread, and try to respond to anyone's flames/comments/questions.
brian...
.