Do any examples of digital SSB modulation exist? If not, is it implausible for some reason? The closest thing I can find is something like 8VSB which comes very close to being single sideband. I read on Wikipedia regarding NTSC (analog VSB) that VSB was chosen over SSB because of issues with attenuation of low baseband frequencies, but I assume this is only an issue for analog SSB modulators and would not be the case if modulation was done in the digital domain as it often is now.
$\begingroup$
$\endgroup$
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
12
Ah, basically everything modern could be understood as single-sideband.
The name is > 100 years old and was based off the fact that traditionally, all you could do on RF is take a single real-valued (and hence, mirror-symmetrical in spectrum) message signal, and mix it up with an oscillator and a leaky mixer, yielding an RF signal that has a carrier, and two symmetrical parts of spectrum below and above.
SSB simply cancels out one of these mirror images – it's not contributing any new info – and typically also the carrier.
Now, modern communication systems simply don't start out from the idea of moving a real-valued signal up in spectrum, preserving and then deleting its inherent symmetry. Instead, our signal modelling assumes a complex, hence typically asymmetric baseband, and moves it up.
So, all the spectra you see from communication devices these days are not symmetric to some carrier. If Richard Carson² was to see these signals' spectra, after reassurance that yes, nothing is broken, we do have that much bandwidth that we need to use, he'd be delighted to recognize that obviously, we've taken his approach of mathematically shaping the signal such that there's no symmetry at the passband very seriously.
¹ there's exceptions to the rule where you want to have a carrier to lock on, but that's not your average terrestrial wireless channel
² who's famous for both "his" rule on how to estimate the bandwidth of an FM transmission, and the invention³ of SSB
³ 1915 US patent; 1917 German patent version, application for which happened 1915-05-22; aside from the technological import of recognizing the orthogonality of $\pi/2$-shifted harmonic oscillations (eq. (6)), I find it historically interesting that the Bell corporation applied for a German patent just two weeks before the US entered the first World War, a good month after the US tore down all diplomatic relations with the Kaiserreich. You'd think that the strong opposition the Wilson administration had against the war Germany waged, especially with Great Britain, would have led to a limitation of how much high-tech and especially communications technology got exported to Germany; even more considering that aside from the humanitarian and geopolitical aversions, the most direct effect of the war on the US at the time was probably naval warfare disrupting US-European trade. And naval warfare definitely depends a lot on wireless communications...
answered Dec 30, 2021 at 12:46
-
$\begingroup$ Thanks for the useful information. I knew that commonly used modulations like QAM are not symmetric, but was curious if at some point past or preset somebody has used discrete SSB for something. I guess the obvious response is "why would they, when QAM and OFDM exists" but honestly I feel like that's just as valid when talking about something like 8VSB which does in fact exist and has been widely used. $\endgroup$– Chris_FCommented Dec 31, 2021 at 4:24
-
$\begingroup$ all this is "discrete" SSB, unless you have a definition of "discrete SSB" that's not just the combination of the usual meaning of "discrete" (i.e., discrete alphabet of symbols in discrete time) and "SSB" (as described above/in the patent). $\endgroup$ Commented Dec 31, 2021 at 11:00
-
$\begingroup$ I suppose I mean specifically something like SSB-ASK. $\endgroup$– Chris_FCommented Dec 31, 2021 at 18:36
-
$\begingroup$ This is the point where I think it's hard to understand what you mean if you don't actually put it into precise words. What is "like SSB-ASK"? What is the defining property that would make something be "like" that, mathematically? Otherwise I can just say "yes" , or "no", because we don't have a common understanding of what you mean. Sadly, you're the only one in a position to clarify that. $\endgroup$ Commented Dec 31, 2021 at 19:00
-
1$\begingroup$ I should have said "modern wireless communications"; ATSC is 30 years old, and as someone of a European school of wireless comms, I'll be as brazen as claiming 8VSB bad idea when it was invented, but heavily lobbied for by the holders of a few of the patents involved. The technical argument for why you'd want to lose the ability to have both I and Q freely available was "making it easier to synchronize by doing a vestigal sideband"; and while the rest of the world actually made the hard thing easier – equalization in frequency domain, America had to way nearly two decades until complicated $\endgroup$ Commented Dec 31, 2021 at 20:35
$\begingroup$
$\endgroup$
There is an example of digital generated ssbonthis github repo https://github.com/threeme3/usdx
