Talk:Software-defined radio

Redirects
It appears that there are three pages for Software Defined Radio. http://en.wikipedia.org/wiki/Software_radio http://en.wikipedia.org/wiki/Software-defined_radio http://en.wikipedia.org/wiki/Software_defined_radio Even though it appears as though they somehow sync-up, I suggest one be maintained and the other two deleted Jennifersteinberg 20:31, 4 June 2007 (UTC)


 * Two of those are redirects to the main Software-defined radio page. Software Receiver is another redirect. Only the main page is being maintained. (Requestion 20:47, 4 June 2007 (UTC))

It isn't 2003 anymore
Current (2003) digital electronics is to slow to process signals from 10kHz to 2.4GHz? Might be a bit outdated. In 2007, I think we're either there or getting there, and while we cannot receive the entire radio spectrum, projects such as HPSDR are developing hardware (the mercury board) that can directly sample the radio spectrum from 0-65MHz. - Ryan 19:26, 15 June 2007 (UTC)
 * Digital hardware is fast enough, but doing it in software is slower. Processors are getting more parallel, so you can do more and more digital signal processing in software, but still slower than in digital hardware. Maybe software-defined radio is the wrong term. Gah4 (talk) 13:38, 11 November 2015 (UTC)

contradiction in definition?
from the intro: SDR [uses] … the soundcard of a general purpose computer (PC), or a reconfigurable home-made piece of digital electronics. But, the military examples given don't sound like either 'a soundcard' or 'home-made digital electronics'. Should it instead say that the processing is done in a type of computer hardware known as a DSP such as what is found in soundcards? 「ѕʀʟ·✎」 06:21, 24 September 2007 (UTC)


 * As SDR is used in the amateur radio market place, any DSP in the soundcard is almost irrelevant to the processing. If there is a DSP there, it is because it is cheaper to base a soundcard on a DSP than to create a logically simpler design.  All the significant processing is actually performed on the main CPU.  The essential function of the soundcard is to do the A/D conversion, and associated anti-alias filtering.  In professional use of the the term SDR, the processing is normally done on a DSP, rather than on a general purpose processor, but that DSP is in the receiver (i.e. it is a complete SDR, not just the front end).  Essentially all the confusion here is because amateur radio mis-uses the terms, so ends up labeling things by their physical realisation, rather than their function. -- David Woolley (talk) 14:04, 22 January 2012 (UTC)

Why "Soundcard"?
In the second paragraph: "An SDR performs significant amounts of signal processing with the soundcard of a general purpose computer (PC)..." Why soundcard? In my opinion soundcard equals to audio card. It's kind of confusing here. —Preceding unsigned comment added by 202.40.139.171 (talk) 08:06, 4 October 2007 (UTC)


 * The quote was nonsense and seems to have been corrected in the main article. The functional component is an A/D converter.  Soundcards are the cheapest way of providing these with a typical PC.  Where they are referred to it is in terms of the physical realisation, not of the the logical structure of the system.  -- David Woolley (talk) 14:16, 22 January 2012 (UTC)

comparsion to oscope/fn-generator
It seems that digital oscilloscopes and function-genertors are bit similar to sdr. any comments? 193.167.107.251 (talk) 19:25, 28 January 2008 (UTC) < This assertion is too vague to respond to.
 * Seems to me that an important part of SDR is the digital mixer. I don't know if that is usual for an oscilloscope, but a digital function generator might make a transmitter, in the same way that an analog function generator makes an AM or FM signal. Gah4 (talk) 19:16, 14 September 2015 (UTC)

Misinterpretation of the Nyquist Theorem
There are several technical inaccuracies in the section on practical receivers. The statement that digital electronics are too slow to receive signals over 40 MHz is incorrect. Remember that the Nyquist Theorem states that the sampling rate must be at least twice the bandwidth of the signal. This means that, in theory, we can sample signals anywhere in the spectrum (with some limitations depending on where the signal is located in relation to the sampling frequency). The limiting factor in actual radios is a measure called the aperture jitter, which quantifies the ability of the ADC to sample at a precise interval (assuming a jitter-free sampling clock, of course). If you are interested more in the subject a google of "undersampling" turns up several excellent references.

The statements on recovering the phase and bit timing are also incorrect. The reason we need to recover phase and bit information is that the transmitter and receiver clocks are not in sync with each other. By recovering this information we will be able to integrate over the entire bit and thus receive the most power and improve the accuracy of the entire radio.

As far as authority goes, I am an electrical engineer who designs digital radios for a living. Roddefig (talk) 07:06, 23 May 2008 (UTC)


 * Part of the idea of SDR is doing as much in the digital side as possible. But yes, if you bandpass filter before the ADC, you are essentially generating a mixer, but in the analog domain. As noted early on, in theory one wants to connect the antenna directly to the ADC, but in practice that doesn't happen. If you generate a programmable bandpass filter, you might be close enough. Gah4 (talk) 19:20, 14 September 2015 (UTC)

Radio protocol?
Can someone define radio protocol, as used several times in the article? --Abdull (talk) 12:49, 24 August 2008 (UTC)


 * Radio protocols are obviously the sets of rules, parameters and conventions needed for radios to communicate. I don't think explicit definition is necessary here. —Preceding unsigned comment added by 24.251.186.204 (talk) 22:25, 23 February 2011 (UTC)

Handset soc chips
The part "Receivers" is copied from NXP itself. There are more companies claiming to have a chip implementation of SDR for a handset soc. from google 217.140.96.21 (talk) 07:51, 3 April 2009 (UTC)

FPGA programming
In the section "SPEAKeasy phase II", at the end, it is claimed that "The time to reprogram these [FPGAs] is an issue limiting application of the radio." This is misleading. The time to write a program for an FPGA is significant; the time to download a stored FPGA program is around 20 milliseconds. This means an SDR could change transmission protocols and frequencies in one fiftieth of a second, probably not an intolerable interruption for that task. I am speaking from my experience with Gigaoperations, the maker of the FPGA computing platform chosen by Xilinx in the 1990s. I imagine with much bigger, but faster, FPGAS today, download times are similar. Rhodesh (talk) 22:43, 10 August 2010 (UTC)


 * Please change the article as you suggest. Regards, Ariconte (talk) 04:12, 11 August 2010 (UTC)


 * Thanks for the change. But 20 milliseconds (ms) are 20/1000 seconds, that is 1/50 second, not 1/5 second. It is true that FPGA got faster, but FPGA got MUCH bigger, too. I think 200 ms as download time for the FPGA configuration is a conservative value. Next, the number of downloads is limited. 1 million downloads before FPGA end-of-live give you only 5.55 hours operation at a configuration change every 20 ms. Question: is 200ms and 1/5 seond correct or 20ms and 1/50 second? AndreAdrian (talk) 14:27, 5 October 2021 (UTC)

The box at the top that complains about insufficient "in-line citations" in 2009
It's half past 2011 at the moment and there are lots of citations. The box should be deleted immediately. It's pure gnome shit.
 * I count about 5 out of 8 sections that have no references, and some of the others are only lightly referenced. WP:V The box is still warranted IMO.--Albany45 (talk) 20:42, 11 July 2011 (UTC)


 * Perhaps it would be better to take into account the NATURE OF THE CONTENT rather than mindlessly demand some unspecified number of citations in each and every section. For instance, READ the sections rather than just count them and then ask yourself exactly what NEEDS referencing.  You may find that the material, considered with its multiple links to wikipedia pages, defines and explains in a self sufficient manner and thus STANDS ALONE without needing any BLEEDING references.  Example: if I say black is sometimes considered white you may sensibly ask for a reference BUT WHEN I SAY FOUR IS 2 + 2, YOU CANNOT, gnome.  — Preceding unsigned comment added by 24.251.186.204 (talk) 20:46, 27 July 2011 (UTC)

Joe Mitola had nothing whatever to do with the 1984 E-Systems software radio
Because of his many publications on the subject, Joseph Mitola III is sometimes called "the father of the software radio." Joe worked, off and on, at E-Systems. In 1984 the phrase "software radio" was defined by members of a team of E-Systems Garland Division employees to describe a large baseband receiver they designed and built. I was a member of that team. Joe was not. We never saw him around and as far as I know he wasn't even cleared to enter the room. He had nothing whatsoever to do with development of the 1984 E-Systems software radio. In the years since then it has annoyed some of us whenever Joe claims he later coined the term software radio (for instance here).

By the way "software defined radio" kind of sucks since it could (and sometimes does) just mean using a GUI to set the knobs, so to speak, on an analog receiver. — Preceding unsigned comment added by 24.251.186.204 (talk) 18:22, 29 August 2011 (UTC)

Today I note somebody (could it have been Joe? ;-) has massively and strangely rewritten the History section. Joe, like you, Midas had nothing to do with the '84 E-Systems radio. We necessarily wrote our own software for all that parallel processing. Garland never used Midas and I don't think we even had it available. When we saw it used at other facilities we considered it awkward, inflexible and, frankly, a little weird. We didn't need no stinking audit trails. — Preceding unsigned comment added by 24.251.169.182 (talk) 14:04, 10 February 2013 (UTC)

Joe, are you saying that a software transceiver is a software radio but a software receiver is not? That would be silly. You wouldn't argue that a hardware receiver isn't a radio. Would you?

The '84 E-Systems Garland Space Systems (hereafter ESY) radio was a surprise to the customer and the relevant community. TTBOMK, ESY at that time was not involved in data analysis. ESY was strictly a system integrator, receiving boxes, designing interfaces and hooking them together. Assemble it, smoke test it, ship it --- that sort of thing. As such, ESY had no need of the Midas analysis software and AFAIK didn't have it.

Now as is well known, the government likes to keep a stable of vendors on call and ready to respond to RFPs. To keep those vendors sweet, projects sometimes seem to be assigned sort of round robin rather than strictly on the merits of the proposals. Anyway, on one particular project, we, ESY, proposed what we called the software radio, which was a great pile of array processors working on the shared memory of an Aptec box, running our unique algorithms effected as data flow programming. This was a front-to-back production engine which necessarily embodied some functions previously performed elsewhere. Tape recordings in, paper out.

Turned out the customer loved the proposal but apparently it wasn't our turn. So somebody else got that contract but the customer gave ESY a special contract to build the radio anyway. And the rest --- say it with me --- was history. My name is John McKown and I approved this message.

Uses Consumer Defnition, not Professional One
The definition of SDR used in this article is basically that used by black box consumers in the amateur radio market. The professional definition of SDR would actually cover nearly all modern, commercially produced, HF amateur radio equipment, and an increasing amount of VHF and UHF equipment.

An SDR is a radio that uses software for back end signal processing, what people are really referring to here are not SDRs, because they actually do not run any software, but rather RF and IF front ends, which can be combined with the A/D converter in a PC sound card, and a general purpose computer, to produce a complete system which is a software defined radio. The definition used here refers to essentially purely analogue devices (although some integrate the A/D converter).

As used here, it is either a marketing term for a specific sub-market, or simply a misunderstanding of the terminology in that market.

-- David Woolley (talk) 12:41, 22 January 2012 (UTC)


 * Huh? No idea what you are talking about. The definition at the top of the article is good enough as it stands. The only improvement needed is to drop the useless word "defined" :-) — Preceding unsigned comment added by 24.251.186.204 (talk) 15:44, 29 March 2012 (UTC)


 * The comment is partly correct, in that "SDR" is a technology that is now used in virtually all but the simplest communications receivers (and transmitters). Any radio that is advertised with "IF DSP" is probably SDR.  The IF is digitized and processed in an embedded computer. It emerges as baseband audio through a DAC.  "Consumers" may not regard such radios as SDR, however. They may think that a separate (visible) computer backend makes it an SDR, or else they see the mysterious terms "I" and "Q". :-)  It would be helpful if the article were clearer about the widespread use of these techniques, even in radios that seem (from the outside) to be conventional.--Albany45 (talk) 02:51, 30 March 2012 (UTC)


 * So he's on about the difference between software and firmware (i.e. can the device owner can reprogram it easily himself)? That's off topic. The scope of the definition and the whole article should just be the digital processing, not hardware implementation details, not tool chains, not amateur vs. professional or whatever.  And, bless your hearts, wikipedia is not well served by sterile editors calling for authors to try again. If you think the definition sucks you should offer a better one yourself. But it doesn't suck.  It's not awesome but it doesn't suck.  — Preceding unsigned comment added by 24.251.186.204 (talk) 16:55, 30 March 2012 (UTC)

Merge in Software GNSS Receiver
I propose that Software GNSS Receiver be merged into Software-defined radio. As an alternative we could just delete Software GNSS Receiver which is more about the technology of hardware vs software radios, than it is about something specific to GPS or GNSS. ★NealMcB★ (talk) 20:16, 24 April 2012 (UTC)


 * Support. I'd go with a merge because the GNSS article has three references. I don't think the article says much, but mention in SDR seems appropriate. Glrx (talk) 19:17, 25 April 2012 (UTC)


 * Meh. I agree with the alternative. Delete Software GNSS Receiver (or not, I don't care.)  Add a sentence or two to Software-defined radio about GNSS radios as examples of SDR -- in the same spirit as you would mention certain cell phones.  The article could use a better section on current application examples.  SDR is not just about Amateur Radio and JTRS.--Albany45 (talk) 02:00, 26 April 2012 (UTC)


 * Just do it :) 2A4Fh56OSA (talk) 08:52, 29 September 2012 (UTC)


 * I suggest keeping the topics seperate. SDR is a general topic whereas GNSS receiver is a specific application of the technology. 92.6.192.31 (talk) 22:50, 26 December 2012 (UTC)


 * Two years have passed, and Software GNSS Receiver has grown, and I don't think you'd want that content here, so I'm removing the merge tag. Fgnievinski (talk) 00:05, 27 June 2014 (UTC)

Separate the SDR Software
I propose to branch the ham radio SDR software to a new article. Furthermore it would be helpful to include a list of the software available on Linux, OSX, Windows as entry points for the interested reader. A list like the revision control software overview would be great. :) 2A4Fh56OSA (talk) 09:21, 29 September 2012 (UTC)


 * This article is losing its way. WP is an encyclopedia. It is WP:NOTHOWTO do stuff. Why is this article recommending a RTL2832U? WP is not a directory of DIY or amateur projects. WP wants its information from reliable secondary sources -- not some self-published website or youtube video. Building an SDR is an accomplishment, but it isn't necessarily notable. I'm willing to break WP rules in many places, but WP should not become a place to advertise projects that have no concrete claim to notability. There need to be WP:RS that indicate a project is significant. Glrx (talk) 19:32, 30 September 2012 (UTC)


 * I think, gnome, that if you actually read that WP guideline on notability you will see that notability is supposed to be a criterion for the existence of a _page_, not its content. Last time I looked, the notability page had a banner or something that said in so many words don't edit content for notability. And wisely so because real people generally note more than gnomes (i.e., volunteer editors for form who imagine they don't need to be knowledgeable about content). — Preceding unsigned comment added by 24.251.169.182 (talk) 16:40, 14 November 2012 (UTC)

Remove all references to sound cards?
Does anybody really consider DSP operations to a baseband signal to be SDR? I would expect that a SDR should be able to encode or decode an AM, FM, or QAM signal by changing the software. If the input to a DSP is via a sound card, then how can the IF frequencies of 455 KHz on AM broadcast and 10.7 MHz on FM broadcast be captured? Does one mix down to a 20 KHz IF? Is it really practical to generate a 20 KHz IF that a sound card could digitize? And could any useful info be extracted from that low, low IF?

I suspect anything that uses the A/D on a sound card is merely a fancy tone control for a baseband signal. That stuff belongs in Audio filter. RastaKins (talk) 21:49, 26 October 2013 (UTC)


 * I'm for removing a lot of the how-to information; does the article really care about specific devices. That said, it is possible to use a 10-kHz IF and decode many modulation schemes. Sound cards are a popular amateur technique to digitally decode many signals. Glrx (talk) 22:53, 27 October 2013 (UTC)

Programmable hardware
Seems to me that the definition of SDR excludes the use of programmable hardware (FPGAs). Maybe this is an accident due to the history of the path leading to current technology. The important distinction should be doing radio operations, such as mixers and filters, in the programmable digital domain, not software vs. hardware. Under the ideal assumptions, you should be able to connect an ADC to an FPGA, and demodulate most radio signals up to hundreds of megahertz, with all the SDR advantages but no software. (In actual implementations, you probably need some filters, though.) Why the restriction to software? Gah4 (talk) 19:38, 14 September 2015 (UTC)

Addition of lots of "marketing-like" material without references
I come back to this article every few months, and lately it seems like there is more and more marketing copy in the text. A lot of it is fairly conspicuous and lacks references (or when references are provided, they are to product / company pages).

Some examples:

Beginning of the History Section:
 * The term "digital receiver" was coined in 1970 by a researcher at a United States Department of Defense laboratory. A laboratory called the Gold Room at TRW in California created a software baseband analysis tool called Midas, which had its operation defined in software.

Basically the entirety of the "Current Usage - Military - USA" section. Here's a portion:
 * ''This goal is achieved through the use of SDR systems based on an internationally endorsed open Software Communications Architecture (SCA). This standard uses CORBA on POSIX operating systems to coordinate various software modules.

The program is providing a flexible new approach to meet diverse soldier communications needs through software programmable radio technology. All functionality and expandability is built upon the SCA.''

“SDRs flexibility results in expensive complexity, inability to optimize, slower ability to apply the latest technology, and rarely a tactical user need (since all users must pick and stay with the one, same radio if they're to communicate).” — Preceding unsigned comment added by 2600:6C64:507F:FFF2:D859:C05A:8AE0:1746 (talk) 21:11, 7 March 2021 (UTC)

The later "Amateur and home use" section also has a lot of stuff like this, where it calls out specific examples of products from SDR vendors, which is a very large market


 * There is a broad range of hardware solutions for radio amateurs and home use. There are professional-grade transceiver solutions, e.g. the Zeus ZS-1[17][18] or the Flex Radio,[19] home-brew solutions,e.g. PicAStar transceiver, the SoftRock SDR kit,[20] and starter or professional receiver solutions, e.g. the FiFi SDR[21] for shortwave, or the Quadrus coherent multi-channel SDR receiver[22] for short wave or VHF/UHF in direct digital mode of operation.


 * WebSDR[28] is a project initiated by Pieter-Tjerk de Boer providing access via browser to multiple SDR receivers worldwide covering the complete shortwave spectrum. Recently he has analyzed Chirp Transmitter signals using the coupled system of receivers.[29]

I'd like to point out that I actually run the GNU Radio project, also listed in this section, and I think the way it is used here is poor form for a Wikipedia page.

The long & short of it is that I think this page needs significant clean-up, but perhaps I'm alone in that opinion.

Also, tangentially, why this article is under the Multiplexing category on Wikipedia is rather confusing to me. — Preceding unsigned comment added by Bhilburn (talk • contribs) 22:54, 6 November 2017 (UTC)

dynamic range?
The article says: ''Real analog-to-digital converters lack the dynamic range to pick up sub-microvolt, nanowatt-power radio signals. Therefore, a low-noise amplifier must precede the conversion step and this device introduces its own problems. For example, if spurious signals are present (which is typical), these compete with the desired signals within the amplifier's dynamic range. They may introduce distortion in the desired signals, or may block them completely.'' It seems to me that this is mixing dynamic range and signal level. To start, microvolt and nanowatt means that the impedance is in milliohms, which is pretty low. One can build 24 bit ADC, though maybe not at the higher speeds. Building an amplifier with noise that low is difficult, and maybe impossible. But the ability to pick up microvolt/nanowatt signals is often due to high power signals, such as nearby transmitters. Without that, an appropriate amplifier isn't so hard to build. Gah4 (talk) 04:41, 1 April 2018 (UTC)


 * Indeed, it's a mess. The concept of spurious signals introducing distortion in the desired signals is probably about strong co-present signals inducing spurious signals via distortion.  These problems are harder in an ADC that's trying to digitize a wide band without a tuned front end to first reduce the co-present signals that are out of band. If you can make that ADC have enough bits and enough linearity and enough speed, then no problem.  In practice, these combinations can often be challenging. Dicklyon (talk) 05:25, 1 April 2018 (UTC)


 * Yes. Good ideas are there, but they seem to mix things up that don't need to be mixed up, besides the voltage and power being wrong. And yes a strong signal causes problems with weaker signals, and yes it is hard to build good enough amplifiers. But the explanations don't seem quite right. Gah4 (talk) 06:47, 1 April 2018 (UTC)


 * Reminds me, I had the idea some time ago about digitizing the whole AM band, which would allow one to change stations, and then back up to the beginning of a program or song. S/N does complicated it if you are near a high powered station. Otherwise, it might work, and the data rate isn't all that high.  A little harder to do the whole FM band, but maybe not impossible. Gah4 (talk) 06:47, 1 April 2018 (UTC)

SDR over IP for reception of AM Digital Radio Mondiale+
Since Wikipedia is for promotion of progress in science and technology.

I would dare to ask, can someone know about mod/hack needed to make such device:

telestar. de / produkt / digiporty-t2 /

into SDR over IP. It does not need to be this exact product, but It need to be stand alone broadcast receiver over IP.

Forgive me for giving description of link, but my English is too poor to describe it in other way, and the Idea is foreign enough for general population.

I am searching for such solution in order to be able to receive AM digital Radio over IP, DLNA.

And I suppose only products of this type/segment can have potential to be cheap enough receiver in order to boost digitalization of radio, and Save Digital Radio Mondiale+ standard from dying out and other standard of digital broadcast over AM.

I have seen so far only raspberry pi with usb dvb-t tuner pluged in used as drm+ over IP server.

But solution seemed amatoure and elaborate.

I do not think people in poor countries could have knowledge how to setup.

And I suppose such type of device have potential to become the cheapest way to receive am digital radio mondiale+ and even dab+.

And I only humbly suppose all mod to achieve desired reception would be to install alternative software on the phone. — Preceding unsigned comment added by 109.173.159.109 (talk) 20:38, 25 August 2018 (UTC)


 * I am a little unsure what you are asking about. Yes radio stations sometimes (maybe often) have an IP streaming system, but it streams the audio, not the RF signal.  This article is about radio receiver, especially the mixer needed to extract audio from RF. That is completely unrelated to IP based radio. Gah4 (talk) 22:22, 25 August 2018 (UTC)


 * Well, not completely unrelated, as IP based radio could go over WiFi or cellular radio signals, but that is still unrelated to the IP part. Gah4 (talk) 22:22, 25 August 2018 (UTC)


 * If we could change/modify such deveice designed to make avaible over IP DVB-t2 signals, just by changing firmware info SDR over IP (exactly as was discovered by DVB-t1 dongles, but they were USB)
 * We could on client side (which could be just mobile phone, android tv box) use just software to listen to DRM+ DAB+ Radio, this Is what I suppose.
 * Because changing firmware in order to get working DLNA server as getway between FM, AM DRM+, DAB+ is more complicated.
 * And I suppose, first step is changing such device into SDR over IP, then in case of success we could think of changing it into DLNA server.
 * However changing it into just SDR over IP is big success.
 * I know you can just plug SDR usb dongle into mobile phone, but it completely DEmobiles it.
 * With very cheap SDR over IP there is only need to run appropriate software in mobile phone. No hanging USB stick, no need to sacrifice expensive good in form of additional smartphone in poor country
 * And that would enable even people in poor countries to listen digital Radio Mondiale on cheap sdr over ip, on cheap smartphone (But, the only one they have) in rural area without internet subscipiton fees.
 * Such Digital Radio from smartphone through bluetooth could be played out in standalone bluetooh speakers.
 * And if we could change it into customizable DLNA service even people in reach countries could use it with comfort.
 * I know you think that you can stick SDR usb into raspberry pi.
 * But raspberry pi cost in poor countries more than cheap smartphone they use.
 * If they were just SDR over IP plug just for frequencies of Digital Radio of different standards {And that made from plug that enables digital terrestrial tv over ip} that would boost digital AM radio even in rural areas with no coverage of digital terrestial tv.
 * I suppose such SDR over IP should cost bellow $25 in order for that to happen. (I have seen such dvb-t1 over IP in similar price but in refurbished form).
 * And Since small revolution came out of USB DVB-T dongle turned into revolutionary cheaply SDR device.
 * I think much greater revolution can possible came out from SDR over IP Ethernet/wifi plug that would be revolutionary cheaply and giving revolutionary possibility without sacrificing additional smartphone Made from DVB-T2 over IP plug.
 * The Good which is not cheap in poor countires.
 * All standalone DRM+ radios are consider even in my country expensive. They cost more than 4k blu ray device which in my country is considered luxury, and there are countries much poorer than mine.
 * My country would be considered pretty rich among poor countries.
 * That is what is hindering Digitization of AM radios, because, people in poor countries, the only one which would be really interested in listening AM radio, can not afford buying Digital Standalone or mobile AM radio equipment.
 * Explanation of purchasing power differance:
 * And people in rich countries cannot even begin to comprehend how big is difference between them and rural areas in the middle of rural Russia, rural Africa {And russia is not even that poor}.
 * In my country I have for $6/couple of years state free mobile internet 512kbps {only one per person} and commercial 32kbps for $1+VATax a year.
 * That makes digital non internet AM radio not very needed. However in rural areas of poor countries the only internet is satellite internet, and check the speed and prices, and multiply it by 16+ that is how different it is.
 * When you think of poor country multiply american prices by 4x and pensions divide by 2. That gives 8x less purchasing power.
 * Then there are 2 magnitudes of poorer countries, and then are rural area in such countries. that is 8x8x8x8 less purchasing power most people can not imagine, because it is fringe of business, existence, and commercial interest. This people have food for free and they are living almost for free. It is not that they do not technology, they are few generation behind newest technology, but newest technology, developed for them can not wait 1.5 decade just to see profit.
 * DRM+ AM radio recievers need to be that cheap from the begining in order to gain their target audience.
 * Purchasing power is like that
 * new 4k BluRay | -> used 4k BluRay -> new BluRay | -> used BluRay -> new DVD | - used DVD / cheapest DVD you can get,  {And DRM+ AM radio need to designed into that area of pricing from the beginning... And I suppose SDR over IP for less than $25 could make it happen.}
 * Of course if such DRM+ AM radio enabling device would be that cheap, even some people in very rich countries would be interested in.
 * And that device in rich countries would be a "little" more expensive. — Preceding unsigned comment added by 109.173.159.109 (talk) 11:42, 26 August 2018 (UTC)


 * SDR and IP radio are two different things. Both, though, use digitized audio compressed in some way.  Most IP radio is not designed to minimize bit-rate enough to satisfy expensive IP links.  Yes it is possible to make in IP programmable SDR that might be able to receive some signals that are otherwise not receivable. But really, one needs to get IP, even at lower bit rates, into more remote areas at affordable prices.  But that isn't the subject of this article. Gah4 (talk) 11:27, 11 December 2019 (UTC)

RF CMOS, and discussion
I reverted a bunch of edits by User:Maestro2016 mostly because I believe that they need discussion before being added. They have also recently been added to, and I believe reverted, from History of radio. I believe that History of radio could use some updating for modern radio systems, though that isn't the question here. I do have some questions for what does and doesn't belong here, and maybe some of the ones I revert really belong. There is a big difference between the ideal SDR and what one really can do, and some existing ICs could be very useful. Lets do some discussing and see what is reasonable. Gah4 (talk) 11:34, 11 December 2019 (UTC)


 * The sources I added mention that modern software-defined radio (late 2000s onwards) is based on RF CMOS technology.
 * You can see the sources for yourself to confirm. Can you be more specific about what issues you have it? Maestro2016 (talk) 18:49, 11 December 2019 (UTC)
 * You can see the sources for yourself to confirm. Can you be more specific about what issues you have it? Maestro2016 (talk) 18:49, 11 December 2019 (UTC)
 * You can see the sources for yourself to confirm. Can you be more specific about what issues you have it? Maestro2016 (talk) 18:49, 11 December 2019 (UTC)
 * You can see the sources for yourself to confirm. Can you be more specific about what issues you have it? Maestro2016 (talk) 18:49, 11 December 2019 (UTC)


 * I was hoping that some others would join the discussion. I think the actual question is how to fit it in with the rest of the article. Most of the article is at a fairly high level, and then getting down to the transistor level without enough in between seems strange. But the whole article is a little strange, as different people have different ideas as to what qualifies.  No processors are fast enough to do a complete radio in the GHz range completely in software. Maybe in the MHz range, so that people can play around with them. Can we wait a few days to see if anyone else has anything to say?  Gah4 (talk) 00:18, 12 December 2019 (UTC)


 * RF CMOS isn't really transistor level, but more IC chip level. It's a type of RF IC chip, rather than a type of transistor. RF CMOS is important in the sense that it enables a SDR system to be downscaled onto a system-on-a-chip (SoC). Either way, the earliest RF CMOS SDR system-on-a-chip I'm aware of is the Broadcom BCM21551 processor from 2007. It's briefly discussed in the Leenaerts source I posted, as well as this source here. So instead of focusing on RF CMOS, we could instead focus more on the BCM21551 processor, with a brief mention of RF CMOS. What do you think about this idea? Maestro2016 (talk) 01:29, 12 December 2019 (UTC)
 * It is transistor specfic as it excludes RF bipolar. Gah4 (talk) 02:29, 12 December 2019 (UTC)
 * Just adding my 2 cents--I think focusing on the processor is the way to go--it is the processors that make SDRs practical, more directly than the underlying semiconductor architecture. -- 01:43, 12 December 2019 (UTC)


 * Part of the problem is that much of the work is likely classified military work, so we don't know the details. It would be nice to have a completely programmable radio receiver that could, with appropriate software, receive anything from WWVB though DirecTV. (To randomly select two). (For those counting, that is 60kHz to 40GHz.)  Antenna-ADC-DSP-output.  That isn't possible. With some programmable analog IC circuitry, maybe one can come close. So, things like programmable RF bandpass filters and programmable IF amplifiers would help. Should the article discuss what exists in those cases?  Gah4 (talk) 02:29, 12 December 2019 (UTC)


 * I've went ahead with the suggested addition. Let me know what you guys think. Maestro2016 (talk) 03:17, 12 December 2019 (UTC)

analog
It seems to me that, as described, SDR is for digital receivers of analog radio signals. That is, traditional modulation systems such as AM, FM, PM, or SSB. But more and more, radio systems are going digital. In other words, there is no ADC in the receiver, but most often some type of spread spectrum system. Programmable digital receivers are now common in, for example, cell phones which need to be able to be used with different cell systems. Only in the case where one needs a receiver for an existing system, such as broadcast AM or FM, or ones such as marine or aviation radio. As I understand it, aviation radio uses AM, as it allows for emergency override, where when two are transmitting on the same frequency, the receiver gets some of each. I suspect, though, that for military applications, they are going digital more and more. Gah4 (talk) 23:14, 3 March 2020 (UTC)

KiwiSDR
KiwiSDR is similar to WebSDR (which can use any SDR radio) but it uses a specific piece of hardware that allows individual 0-30 MHz tuning for each user. Link to website: http://www.kiwisdr.com/ — Preceding unsigned comment added by 194.193.212.24 (talk) 05:16, 17 October 2020 (UTC)

GS1299
There is a very interesting GS1299 single chip FM radio receiver. The data sheet doesn't give much details, but it goes from antenna with a fixed LC filter on one side, to stereo audio outputs, able to drive headphones, on the other side. The only tuned circuit is a 32768Hz crystal oscillator. I don't see any mention or reason for software, though. It does say CMOS, and much of the processing can be done in digital CMOS hardware without any software. It is fixed for the FM band, so no software programming needed. As far as I know, though, there is no article on non-software digital receivers for analog radio signals. In any case, I assembled a kit with this chip, and it works just fine. Gah4 (talk) 02:18, 24 February 2021 (UTC)

Waveforms?
Text excerpt: "widely different radio protocols (sometimes referred to as waveforms)" Yrs trly is an old–timer; that ref. to waveforms looks strikingly peculiar. If it has a special meaning within the context of SDR, please clarify. Nevertheless, especially after watching two videos about specific SDRs, I know that I want to learn more! Nikevich 05:59, 1 December 2021 (UTC)
 * It is military jargon, I think--waveform refers to the software setup of an SDR, and thus the radio protocol, for example, and . --  06:46, 1 December 2021 (UTC)
 * Waveform is the general term for a time-domain analog signal, especially if it can be viewed on an oscilloscope. It is mostly useful when some characteristics, such as amplitude or phase, are ignored. It can be used for digitally generated analog signals, but not, as well as I know, for digital signals themself. One can describe the character of an FM radio signal in general, or usually in terms of modulation index. Gah4 (talk) 07:39, 1 December 2021 (UTC)
 * Waveform is the general term for a time-domain analog signal, especially if it can be viewed on an oscilloscope. It is mostly useful when some characteristics, such as amplitude or phase, are ignored. It can be used for digitally generated analog signals, but not, as well as I know, for digital signals themself. One can describe the character of an FM radio signal in general, or usually in terms of modulation index. Gah4 (talk) 07:39, 1 December 2021 (UTC)
 * Waveform is the general term for a time-domain analog signal, especially if it can be viewed on an oscilloscope. It is mostly useful when some characteristics, such as amplitude or phase, are ignored. It can be used for digitally generated analog signals, but not, as well as I know, for digital signals themself. One can describe the character of an FM radio signal in general, or usually in terms of modulation index. Gah4 (talk) 07:39, 1 December 2021 (UTC)

Complete USRP section is gibberish
Hi. "The" GNU Radio: I was the head maintainer for GNU Radio, nobody uses it with the "the" article; it's a software project, just like "the libreoffice".

I also been a Support Engineer for Ettus for nearly a decade. There's several separate USRP models, but none correctly match the technical data given in the section.

The only forgivable sin here is that GNU Radio used to be predominantly used with USRP hardware. That isn't as strongly the case anymore, since better integration of other SDR hardware has arrived.

I'm hence reworking that section. Marcusmueller ettus (talk) 12:09, 3 January 2023 (UTC)

Introductory section meanders, badly
There's absolutely no need to bring a bullet list of arbitrarily chosen waveforms to say "SDR enables usage of different waveforms"; it actively makes the article worse.

I'm removing that list, it has no business being in this article. A link to Modulation serves the same purpose, just better! Marcusmueller ettus (talk) 12:24, 3 January 2023 (UTC)