Re: [Discuss-gnuradio] Wideband Random Noise Cypherpunk Guerrilla Radio - Doc Req

[Marcus Müller]

Hi Grarpamp,
On 01.06.2016 05:24, grarpamp wrote:
> On 5/29/16, Marcus Müller <address@hidden> wrote:
>> If you spread it extremely wide and basically put the power level
> Probably more like this than in your first larger paragraph.
Well, then this discussion gets quickly out of scope of GNU Radio – you
see, GNU Radio is mostly for complex baseband signal processing, and
that kind of radio really depends on things being done in hardware. So,
not SDR.
> Perhaps crafting arbitrarily fine random vertical divisions of a
> spectrum into a very wide line of numerous bit positions, such lines
> being known only to peers. To which they may apply essentially gain
> hits ("power"), perhaps in parallel, to carry data. Looks like
> random background noise to outside observers.
There's many things bad about that approach: What you describe is a
Q-Tone FSK (I think that was the official name; it gets pretty different
names from different people); it's well-researched, data capacity wise
worse than FSK scaled to achieve the same bit rate at the same SNR, and
most of all: it's very well-detectable by parametric spectrum
estimation. If you want to look into that: ESPRIT with a broadband
signal pre-suppressor would be the go-to algorithm if I know I'm looking
for a set of tones.

Also, if those tones need to "look" very narrow, they can, by math, not
be modulated fast (because that would convolve the individual's tone
spectrum with the symbol shape, and hence decrease narrowness).
> The paper may have even been posted here, but without an unaltered raw
> copy of the list archives in mbox format available for download for
> people to then load into local MUA's, it's really hard to search for
> such text, links, and attachments. Such an archive would be highly
> appreciated and useful :) 
Such a paper hasn't been posted here. Generally, we (sadly, maybe) don't
see that many people post their papers here – then again, they should
either way only do so on the discuss-gnuradio list if their research
included GR in the first place. But then again, if everyone who's
written an IEEE paper containing the word "GNU Radio" in the metadata,
we'd have 322 of those [1] and for "GNU Radio" *anywhere* in the
article, we'd have 2443 [2] (today). I don't think we could fully
appreciate that level of sharing. I think that number, 2443, highlights
how much research and publication there has been over the last few
decades on SDR. There's tens, if not hundreds, of thousand of papers
that deal with theoretical and/or practical research on SDR systems.

So, I'd really like to explain that both secrecy and privacy have been
important, and fruitful, research topic for radio researchers of the
last 70 years; there's really nothing "new" about inventing a waveform
that distributes  Bit Energy over a larger spectral area.

So, frankly, I think this discussion with my walls of text and your
hopeful investigations is terribly off-topic here :( as it has not the
least to do with GNU Radio. Also, we're not a library of papers.

The main problem I see is that it's probably going to be off-topic
everywhere, because frankly, you don't seem to have too much of a clue
about spread-spectrum digital communications; so I'd really encourage
you to use your enthusiasm to humanities (or your own) best! In my first
email, I used some termini technici to describe what you were describing.
Your goal should probably be understanding DSSS and CDMA. That won't
happen by just reading the wikipedia articles on those, as you'll need
first get your digital comms knowledge about non-spread spectrum theory
straight.
Then, you should probably read a bit up on information theory – just so
you get a grip on what Shannon's Entropy is, and then move on to
understanding "physical layer privacy" (which, in fact, seems to be a
hot research topic right now, or for the last 2 years, whatever);
admittedly, the is a bit of a tough one, and I'm not fully convinced
research has yet led somewhere interesting (and even more so, I think
some of the specific research that has been done and funded leads
exactly nowhere, but that's a rant for another day).

With this, I'd like to ask you to wrap this discussion up on this
mailing list: it's really been a lot of text, but I'm afraid, I'm not
doing the others on the mailing list a favor by keeping to reply to you.
Also, I've got stuff to do :)

So: get your digital comms textbooks ready, and use GNU Radio in the
process of learning practical comms – it's really a great tool for the
case that you really just want to look at what the spectrum of something
that you can also easily create with GNU Radio. Please DO come back here
and discuss problems you encountered with GNU Radio on your journey! As
you might have noticed by the pages worth of text you're getting as
answers, we do enjoy a good discussion on-topic and quite a bit around
topic :)

Best regards,
Marcus


[1]
http://ieeexplore.ieee.org/search/searchresult.jsp?action=search&sortType=&rowsPerPage=&searchField=Search_All&matchBoolean=true&queryText=%28GNU%20Radio%29
[2]
http://ieeexplore.ieee.org/search/searchresult.jsp?action=search&sortType=&rowsPerPage=&searchField=Search_All_Text&matchBoolean=true&queryText=%28GNU%20Radio%29