Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?
Get HideMyAss! VPN, PC Mag's Top 10 VPNs of 2016 for 55% off for a Limited Time ×
Communications GNU is Not Unix Open Source Hardware

Getting Started With GNU Radio ( 42

An anonymous reader writes: Software Defined Radio must be hard to create, right? Tools like GNU Radio and GNU Radio Companion make it much easier to build radios that can tune AM, FM, and even many digital modes. Of course, you need some kind of radio hardware, right? Not exactly. Hackaday has one of their video hands on tutorials about how to use GNU Radio with no extra hardware (or, optionally, a sound card that you probably already have). The catch? Well, you can't do real radio that way, but you can learn the basics and do audio DSP. The next installment promises to use some real SDR hardware and build an actual radio. But if you ever wanted to see if it was worth buying SDR hardware, this is a good way to see how you like working with GNU Radio before you spend any money.
This discussion has been archived. No new comments can be posted.

Getting Started With GNU Radio

Comments Filter:
  • SDR Hardware (Score:5, Insightful)

    by PPH ( 736903 ) on Thursday November 12, 2015 @01:42PM (#50915975)

    Entry level, about $12 []. I think I'll just go ahead and risk it.

    • Re:SDR Hardware (Score:5, Interesting)

      by drinkypoo ( 153816 ) <> on Thursday November 12, 2015 @01:57PM (#50916055) Homepage Journal

      You piqued my curiosity so I plugged the part numbers for the ICs into eBay and got back the same thing for ten bucks shipped []. If I were already doing a newegg order, though, I might well toss one in from them so as to get it this month.

      • You can download my SdrDx [] for either Windows or OS X, download a saved RF file, and start receiving from a recording of, for instance, a ham band during a contest, or a SW band with some interesting stations on it. No SDR required to fool around, and the software is free.

        You can tune around, play with bandwidths, demodulation modes, noise blanking, peak tracking, notch and other filtering, the analysis scope, etc. WIth a recording, you get the span of the spectrum that was recorded (for instance, 200 khz of

    • Re:SDR Hardware (Score:5, Informative)

      by Tailhook ( 98486 ) on Thursday November 12, 2015 @01:58PM (#50916061)

      If you can add $10 to that sell R820T/RTL2832U dongles (on Amazon []) with temperature compensated oscillators, SMA connectors and other nice features for SDR experimentation. Start with that if you imagine using upconverters, front-end filters, etc.

      You'll want a short USB pigtail for these devices, though; they are fairly large.

      • Or you can stabilize the temperature of the crystal yourself. We used to build crystal ovens to stabilize crystal oscillators at a constant temperature, and heat it up to 40C (or 50C depending on where you live) with some resistors. Use a thermistor and a regulator to control the temp.

        Even without the oven, the R820 dongles on ebay from china gives you good stability after 10 minutes. You can listen to FM radio, Aircraft radio, old cellphones. I have one set up to receive aircraft tracking info as well as o

        • by Agripa ( 139780 )

          Crystals designed for operation inside of an oven use a different cut, usually SC, to move their zero temperature coefficient point to a higher temperature. Common AT cut crystals are optimized to operate at 25C so they are not ideal for this. It might be fun though to use a peltier element for both heating and cooling to maintain a 25C temperature.

          When I have hacked together this kind of circuit in the past, I used a TO-220 style transistor as both the heater and sense element with an analog sampled cont

    • Re:SDR Hardware (Score:5, Interesting)

      by lkcl ( 517947 ) <> on Thursday November 12, 2015 @02:00PM (#50916079) Homepage

      yeah i got something similar, i researched the chipset a bit in advance, to see what frequencies it could do. i found one that could go from i think it was... like... 100 mhz through to almost 1900mhz, with a bandwidth of something like... 2.8m-samples/sec. it only had an 8-bit ADC resolution but that was ok. i then used it with some software i was working with, at the time (passive GSM scanning software), and actually managed to find a couple of frequencies, which was amazing.

      going beyond that would have been tricky, because at the limit of 2.8 million samples per second of I and Q data @ 8 bit, it was pushing the limit of what the hardware could actually do: there were quite a few drop-outs. i'm sure the proprietary driver could handle that data rate, but the reverse-engineered gnu/linux one simply couldn't.

      anyway yes absolutely! $12 plus shipping for something that will handle a huge range of frequencies, FM radio, TV frequencies, GPS satellites, GSM 850 and 900mhz, and even some of the higher-end GSM frequencies @ 1800mhz... maan, what more could you ask for? :)

      • by Eythian ( 552130 )

        going beyond that would have been tricky, because at the limit of 2.8 million samples per second of I and Q data @ 8 bit, it was pushing the limit of what the hardware could actually do: there were quite a few drop-outs. i'm sure the proprietary driver could handle that data rate, but the reverse-engineered gnu/linux one simply couldn't.

        2.8msps is a limit of the hardware in the mode that uses them as an SDR. In DVB-T mode, it works somewhat differently and so can do higher bandwidth, but as an SDR it's in a

    • by fuzzyf ( 1129635 )
      It's worth mentioning that Michael Ossmann (creator of HackRF) is recording an online course about SDR and digital signal processing. It's ongoing, so there will be more videos as times go by.
      It mentions hackrf but you can do get by just fine with a regulrar cheap dongle. []

      Thank you Michael Ossmann! :)
  • by fuzzyfuzzyfungus ( 1223518 ) on Thursday November 12, 2015 @02:10PM (#50916145) Journal
    Back in the day, I could blame crippling hardware costs for my ignorance of signal processing. Now what am I going to do?
  • by mrops ( 927562 ) on Thursday November 12, 2015 @02:18PM (#50916203)

    Where I live, radar detectors are illegal.

    Naturally, police have Radar detector-detector built into their radars, and I know of one confirmed case where a friend visiting out of state got caught with his radar detector.

    So, can a radar detector built out of SDR/GNU Radio be detected? Hypothetically of coarse.

    • Yes, it probably will detect it, at least for inexpensive SDR hardware. Most modern receivers are the superheterodyne type, in which an oscillator within the receiver is set to a frequency near the frequency you wish to detect. This simplifies the circuitry and software because you're only processing the DIFFERENCE between the received signal and the reference, rather than directly processing the source waveform at some sample-rate multiple of the frequency of interest.

      The detector-detector picks up the

      • There's no way to shield that form of RF leakage through the antenna without shielding the antenna... which would make it unable to detect anything to begin with. The problem is with the device itself and building a zero-leakage superhet receiver requires quite a better set of components available to consumer electronics makers. Which is exactly why detector-detectors work even against "undetectable" snake-oil devices.

        • >There's no way to shield that form of RF leakage through the antenna without shielding the antenna... which would make it unable to detect anything to begin with.

          Yeah I mean to add it would not be easy, if it's possible at all.

          "There's no way" to do it, much like we used to say "there's no way" to alter a file without changing it's MD5 hash, and we said "there's no way" to install 3rd-party software on an iPhone, etc. Impossible, until some clever person figures out how to do it.

          I was a professional mag

        • Actually it can be done with a little RF theory.

          The way these things is by multiplying the incoming RF by a carrier of frequency very close to the radar frequency. The result is a lower-frequency product that can be sampled or simply compared in the analog domain to a reference. In this setup, the frequency we are multiplying by (the "LO" or "local oscillator") is what is detected remotely, because, as you point out, it has a very direct connection to the antenna.

          Here's how you get around it:

          Multiply by the

        • Well there are ferrite isolators. []
      • by mrops ( 927562 )

        Hmmm, so no building a stealthy detector this way.

        Is it theoretically possible to capture a larger bandwidth using SDR, with oscillator freq far away so not to be detected, then doing a FFT on it to isolate police radar?

        In my excuse if I make no sense, last I studied radio communication was 2 decades ago in engg, went down to software development and never looked back.

        • All you need to do is use a conversion scheme that isn't the same as those used by common radar detectors.

          For instance, it might be common in the detector industry to use a first IF at 7 GHz. This could be convenient because the same 17 GHz LO would be useful for both 24 GHz K band and 10 GHz X band reception. I have no idea if that's what the manufacturers actually do, but if they do, then the "detector detectors" might work by listening for leakage at 17 GHz. Put your first LO in a different part of th

      • by Ozoner ( 1406169 )

        A simple (well shielded) RF stage is all that's needed to block Local Oscillator radiation from the Antenna.

        Plus the choice of IF stage decides what frequency the L.O. will be on, so false detection's will make any "radar detector, detector" pretty much useless.

        The existing Radar Detectors rely on consumer grade Radar Detectors being very primitive devices with a passive detector (eg no shielding)

      • by Agripa ( 139780 )

        One of the techniques I used when transmitter hunting was to look for the local oscillator frequency when the transmitter was not active. This was especially useful for low duty cycle transmitters but it was tricky because one antenna and receiver usually will not cover both frequencies well and finding the local oscillator frequency may be difficult.

        Superheterodyne receivers usually include an RF amplifier before the first mixer stage just to attenuate leakage of the local oscillator; without it, meeting

    • iirc the lowest band police radar transmit in is 10ghz, with some of the other bands being as high as 24ghz. The SDR dongles generally have a response frequency of around double digit MHZ to a bit over 2.4 ghz. That's almost an order of magnitude lower than what you'd need to detect a radar gun. From what I know of radar detector detectors, they work based on listening for the detector to emit slight echos whenever its oscillator is oscillated by an incoming beam. As such, I doubt there's any interaction be
    • For the most part, yes, it can be detected.

      That is because when processing the radar signal in order to alert you, it emits another frequency signal and there would be no real way around it without a bunch of larger equipment like antennas and amplifiers or shielding. A radar detector's receiver called a superheterodyne receiver. [] It emits a frequency that interferes with the radar frequency which in turn allows reception on lower frequencies that are easier amplified.

      To get around this, you would need a lar

  • by Rufty ( 37223 ) on Thursday November 12, 2015 @02:45PM (#50916449) Homepage

    For transmitting there's the HackRF [] which is a few mW output and is the one I've played with. Also another supplier that has cheaper, transmit only versions; the HackRF Blue []
    For quite a bit more, there's MIMO capable devices such as the Ettus USRP [] that lets you run your own GSM basestation [] among other things.
    And for a more stand alone device, there's always the PortableSDR []
    I've got a HackRF and am having fun with it trying to make a network analyzer. The others, I've just heard about.

    • by Anonymous Coward

      Worth noting you need an amateur radio operator license in most countries to transmit on most frequencies. On frequencies for which you don't need a license, it has to be done with a licensed device (so licensed because it's locked to a frequency or range and its transmission power is limited to a very short range).

  • can GNURadio-companion incorporate hamlib as a library to run the radio and basically use gnuradio-companion to build some simple front ends for hamlib? i have a Ten-Tec RX320 and Grig is not much as a front end, i was hoping to build a nicer front end that was more specific to this HF receiver and include a waterfall feature to get a good look at a few KCs of spectrum, i use Linux exclussively so there is not a windows PC for me to do this in, if i get good and skilled at this i would love to build a FOSS
  • My favorite SDR platforms. I have a cheap RTL dongle that I use for just about everything outside the ham bands. I use soundmodem along with aprx to run my digipeater/igate. fldigi takes care of PSK and other modes on HF. GNURadio and SDR# for listening to what is going on around town.

    Fun stuff.

  • I too was attracted by the allure of "approachable" radio communication system design....

    I've been working on my Masters thesis involving GNU Radio. I have an RTL-SDR (a Terratec Elonics E4000) and more importantly 2 bladeRF x40 SDRs. Observing/listening/decoding certain transmissions with pre-existing standards is fairly easy. Building a complete digital data transmitter and receiver in GNURadio Companion has a bit of a learning curve. And by "learning curve" I mean "like free climbing the Dawn Wall".

RAM wasn't built in a day.